Evaluation of four airway training manikins as patient simulators for the insertion of single use laryngeal mask airways*


  • *

    Presented in part at Difficult Airway Society Annual Meeting, Dublin, Ireland; 4–6 October 2006.

Dr Tim Cook E-mail: timcook007@googlemail.com


We evaluated the performance of four manikins: Airsim, Bill 1, Airway Management TrainerTM and Airway TrainerTM, as simulators for insertion of single-use laryngeal mask airways and the reusable LMA Classic. Sixteen volunteer anaesthetists inserted each laryngeal mask airway into each manikin twice. Insertions were scored for ease of insertion, clinical and fibreoptic position, and lung ventilation (maximum score 10). Scores < 7 were classified ‘poor’ and < 5 ‘failure’. We analysed manikin and laryngeal mask airway performance. Poor insertion rate was 15% (range 9–21%) and was lowest for the VBM manikin (p = 0.02). Insertion failure rate was 2.6% and did not differ significantly between manikins (p = 0.2). Overall manikin performance was significantly different (p < 0.0001). The VBM manikin scored best, with all other manikins equivalent. The VBM manikin performed significantly better for three individual laryngeal mask airways. Overall performance differences of laryngeal mask airways were statistically significant (p < 0.001) but individual comparisons were not. Silicone devices performed better than PVC devices (p < 0.05) Devices with and without grilles performed similarly. All manikins were adequate. The VBM manikin performed best overall and for several individual laryngeal mask airways. The methodology is useful for future evaluations of devices, both manikins and supraglottic airways. Further human clinical research is required.

A large number of laryngeal mask airways are now available to compete with the original LMA Classic (Intavent Orthofix, Maidenhead, UK) for use in anaesthesia and cardiopulmonary resuscitation [1, 2]. Most of these laryngeal mask airways are single-use devices, with most made of polyvinylchloride (PVC), though a few are made of silicone. Manikins are routinely used for practical training in the use of laryngeal mask airways and may also be used for studying the performance of new devices. Manikins have traditionally been designed for practising facemask ventilation and or tracheal intubation, but are frequently used with the LMA Classic and more recently other laryngeal mask airways. It is not know whether currently available manikins perform acceptably with these devices, which manikin is best, or whether different manikins perform equally well for different laryngeal mask airways. Finally there are few comparative studies of these laryngeal mask airways [3]. This study sets out to answer these questions using recently developed manikins designed for use with supraglottic airway devices including laryngeal mask airways.

The study aims were to determine the following:

  • • which manikin performed best for all single use laryngeal mask airways;

  • • which manikin is best for each individual laryngeal mask airway;

  • • the performance of individual laryngeal mask airways in all the manikins.


The study was approved by the Local Research and Ethics Committee and each voluntary participant gave written consent.

We evaluated the following manikins:

  • • the Airway Management Trainer (Ambu, St Ives, UK);

  • • the Airway Trainer (Laerdal, Stavanger, Norway);

  • • Airsim (Trucorp, Belfast, Northern Ireland);

  • • ‘Bill 1’ (VBM GmbH, Sulz, Germany).

These are referred to by their manufacturer (i.e. Ambu, Laerdal, Trucorp and VBM). Each manikin studied was designed for laryngeal mask airway insertion and was the latest version available in March 2006 (Fig. 1, Table 1).

Figure 1.

 Airway manikins studied.

Table 1.   Manikins.
  Airway TrainerAirway Management Trainer‘Bill 1’ Airway
Management Simulator
  • *

    Resusci-Anne head released as prototype.

  • For head to attach to existing Resusci-Anne manikin/for head with electronic sensors for use with recording Resusci-Anne manikin.

Date releasedDecember 2004*February 2005December 2004March 2005
ManufacturerLaerdal Medical,
 Stavanger, Norway
Ambu Ltd. Burrel Road,
 St Ives, Cambridgeshire
 PE27 3LE
VBM Medizintecknik GmbH,
 Einsteinstrasse 1, D-72172
 Sulz a.N., Germany
Trucorp Ltd, Whitla Medical
 Building, 97 Lisburn Road,
 Belfast BT9 6JQ
Approximate cost;
 £ (€)
600 (871)/800 (1162)600 (871)1250 (1816)1200 (1743)

The following laryngeal mask airways were studied:

Ambu LM (Ambu, St Ives, Cambridgeshire, UK), Laryngeal airway device (LAD)* (Marshall Medical, Bath, Banes, UK), Softseal (Smiths Medical, Hythe, UK), Solus (Intersurgical, Wokingham, UK), LMA Unique (Intavent-Orthofix), Proact laryngeal mask airway* (Proact, Great Oakley, UK), LMA Classic* (Intavent Orthofix). All are single-use devices except the LMA Classic, which was included as a comparator. Those laryngeal mask airways marked with an asterisk are made of silicone, those without are made of PVC. All devices were those currently marketed by the manufacturers in July 2006.

For each insertion of a laryngeal mask airway, a lubricant (of a type recommended by that manikin's manufacturer) was used both in the manikin and on the posterior of the laryngeal mask airway.

Prior to the study, sizes 3–5 of each laryngeal mask airway were inserted in each manikin six times by the investigators and scored according to the test protocol (see below), to determine for each laryngeal mask airway the ‘best size’ for each manikin. In most cases the size 4 scored highest, though differences between this and size 3 were generally small. This data was not included in the analysis of device performance. The ‘best sizes’ of each laryngeal mask airway for each manikin are listed in Table 2.

Table 2.   Sizes of laryngeal mask airways used in the four manikins: ‘best size’.
Intavent LMA Classic4444
Intavent LMA Unique4444

Sixteen volunteer anaesthetists inserted the ‘best size’ laryngeal mask airway into each manikin in random order, twice. The order of manikin and laryngeal mask airway use was randomised by picking names from a hat. All volunteers had prior experience with the LMA Classic but not with the single use laryngeal mask airways. Each insertion was scored for ease of insertion (2 = easy, 1 = difficult, 0 = not possible), clinical positioning (presence of an ‘end point’ to insertion, mask rising out with inflation, laryngeal mask airway tube remaining in midline all scoring 1), fibreoptic position (2 = vocal cords visible, 1 = other laryngeal structures visible, 0 = no laryngeal structure visible), ability to ventilate the ‘lungs’ (2 = good ventilation, 1 = poor ventilation, 0 = ventilation not possible) and whether performance was lifelike (1 = lifelike, 0 = not lifelike). Maximum score was 10. A score of < 5 was considered ‘failure’ and < 7 to be ‘poor’. Data were analysed to determine adequacy of performance, explore performance differences between manikins overall, for individual laryngeal mask airways and to explore differences in laryngeal mask airway performance.

Single-use laryngeal mask airways were re-used throughout the study, provided there were no external signs of damage or malfunction. There were no failures of the devices. Five of each laryngeal mask airway were available and each was used approximately 25 times.

Statistical analysis

Friedman's anova with post hoc Wilcoxon signed rank testing was used for repeated ordinal data, the Chi-squared test for summed categorical data, and the Mann–Whitney U-test for non-paired groups. p < 0.05 (corrected for multiple comparisons) was taken as significant.


Manikin performance for all single use laryngeal mask airways

A total of 224 laryngeal mask airway insertions were made in each manikin. Insertion failure was 2.6% overall (range 0.9%– 4%) (p = 0.22) and 15% of insertions were poor (range 9–21%, with VBM best) (p = 0.02). Overall scores between manikins differed significantly (Friedman p < 0.0001) and laryngeal mask airway insertion in the VBM manikin scored highest (p = 0.006 compared to each of the other manikins). Differences between other manikins did not reach statistical significance (p > 0.05) (Table 3).

Table 3.   Performance of the manikins overall: for insertion of all seven laryngeal mask airways.
 Insertion scores
  1. Scores: Friedman anova p < 0.0001. Significant differences (post hoc corrected p < 0.05, > indicates ‘better than’): VBM > Ambu, VBM > Laerdal, VBM > Trucorp.

Median 89 8 8
IQR 7–98–10 7–9 7–9
Mean 7.958.6 7.93 7.88
Score < 7: ‘poor’21%9%12%18%
Score < 5: ‘failure’ 4%2.7% 2.7% 0.9%

Manikin performance for each laryngeal mask airway

Each manikin had each laryngeal mask airway inserted 32 times so the power of the study to detect such differences was low. Manikin performance varied significantly (Friedman p < 0.05) for four of the seven laryngeal mask airways examined, with a p value of < 0.07 for two other laryngeal mask airways. For the Marshall LAD and Proact laryngeal mask airway the VBM manikin scored statistically significantly higher than all other manikins (Friedman anova p < 0.05 and post hoc Mann–Whitney p < 0.05). Insertion of the Intersurgical laryngeal mask airway scored significantly higher in the VBM manikin than either the Laerdal or the Ambu manikins. LMA Unique insertion in the VBM manikin scored significantly higher than in the Laerdal manikin. None of the other comparisons reached statistical significance.

Analysis of laryngeal mask airway performance

Each laryngeal mask airway was inserted 128 times in total. Laryngeal mask airway performance varied significantly (Friedman p < 0.0001). Rates of failed and poor insertion for individual laryngeal mask airways ranged from 1 to 7% and from 10 to 32%, respectively, with the Unique and Intersurgical laryngeal mask airways performing least well in both measurements. The power of the study to detect such differences was low: several differences of statistical significance were recorded but without providing an overall clear ranking. Based on summed scores, the LMA Classic, Marshall LAD and Proact laryngeal mask airway performed best, followed by the Ambu laryngeal mask airway. Poorest performance was by the Intersurgical Solus and LMA Unique (Table 4).

Table 4.   Performance of each laryngeal mask airway. Mean, median and interquartile range (IQR) presented as scores. ‘Poor’ and ‘Failure’ refer to insertion scores < 7 and < 5, respectively. Presented as percentage of insertions.
LMA Classic
LMA Unique
  1. Scores: Friedman anova p < 0.0001. Significant differences (post hoc corrected p < 0.05, > indicates ‘better than’): Ambu > Unique, LMA Classic > LMA Unique, Marshall > LMA Unique, Marshall > Intersurgical, Proact > Unique, Proact > Intersurgical.

IQR 7–97.8–9 7–9 6.8–9 7–9 8–9.3 7–9

Analysis of laryngeal mask airway type and performance

Silicone devices performed statistically significantly better than PVC devices (median score 9 vs 8; p = 0.0003, poor 10% vs 19% p = 0.0002, failure 1.8% vs 3.3%; p = 0.27).

Performance of laryngeal mask airways with grilles (the LMA Classic and LMA Unique, both manufactured by Intavent Orthofix) and without (all others) did not differ significantly (median score 8 vs 8 p = 0.2, poor 16% vs 15% p = 0.89, failure 3.1% vs 2.3% p = 0.66).


The primary goal of our study was to examine manikin performance for laryngeal mask airway insertion. Our results indicate that manikin performance for single-use laryngeal mask airway insertion is unequal, although all four of the manikins we evaluated performed adequately, with the highest failure rate being 4%. For a department wishing to use a single manikin for training with a variety of laryngeal mask airways all manikins are probably adequate but the VBM manikin performed best and the Ambu manikin least well.

It is reasonable to ask whether the manikin with a higher score in this assessment is a better simulator of laryngeal mask airway insertion than manikins with lower scores. When using the correct technique in vivo, most laryngeal mask airways are inserted with success rates exceeding 85% and often around 95%. Laryngeal mask airway insertion in manikins is often more difficult and fails more often. Difficulty with laryngeal mask airway insertion (and ventilation) is likely to be frustrating for candidates and trainers and is not ‘lifelike’. Manikins that also simulate additional features of laryngeal mask airway use, such as clinical tests of correct positioning, are more useful than those that do not. Our scoring system was designed to assess all those measures that are important in teaching laryngeal mask airway insertion: it includes insertion and ventilation success, but also incorporates measures of clinical and anatomical position and whether performance was lifelike. For this reason we believe that a higher scoring manikin does represent ‘better simulation’ of the overall features of laryngeal mask airway use and will be of greater utility as a simulator. In this study using correct insertion techniques we found ‘poor’ insertion rates of 10–32%; these ‘poor’ scores (i.e. < 7/10) probably represent insertions that would be considered unacceptable in vivo. This implies that laryngeal mask airway insertion in all these manikins is probably not as easy as in human subjects and this would fit with our clinical observations. Overall, we suggest that higher scores in this study are likely to identify manikins that are the most useful and appropriate for training laryngeal mask airway insertion.

We have also been able to examine the performance of each manikin for individual laryngeal mask airways. This has allowed us to determine which manikin is ‘most suited’ to a particular laryngeal mask airway. In spite of the rather limited power of this study to detect these differences, statistically significant differences in manikin performance were found for insertion of four of the seven laryngeal mask airways used. In each case the VBM manikin was a high performer. Further work may be warranted before choosing a particular manikin to evaluate or train with a particular laryngeal mask airway. Data on individual laryngeal mask airways may also be useful when designing and interpreting manikin studies.

The study methodology has also enabled us to examine the relative performance of each laryngeal mask airway. Again, we found significant differences in overall performance. The study had limited power to differentiate between individual masks. With seven laryngeal mask airways studied, each p value requires a 21-fold correction factor, making detection of statistically significant differences a challenge. The results do not allow a clear ‘ranking’ of the mask airways but several silicone mask airways score notably better than several PVC masks.

This observation was confirmed when we compared the relative performance of silicone and PVC masks and of masks with and without grilles. Although we found no difference in performance of laryngeal mask airways with and without grilles, we did find that the laryngeal mask airways made of silicone performed significantly better than those made of PVC. There are surprisingly few published comparative evaluations of the laryngeal mask airways, with no published data on four of the devices we studied. These data might stimulate further clinical evaluation.

There are several limitations to our study. Firstly, although we have used the most up-to-date versions of the manikins available at the time of the study, modifications may have been made subsequently by the manufacturers. Secondly, we have evaluated only one of each manikin: there may be variations between individual manikins of the same type despite originating from the same production line. Similar arguments of consistency of design and performance also apply to the laryngeal mask airways used, though we are not aware of changes to any laryngeal mask airways since the study was performed. We have not examined consistency of manikin performance over a prolonged period or durability of the manikins. Also, despite concluding that the VBM manikin performed better than the other manikins for insertion of these laryngeal mask airways, we have not compared cost (Table 1) or other features or performance qualities which may affect a department's choice. Finally, we have compared how these devices perform with each manikin. This does not equate to determining how each device performs in human subjects. Clinical observation suggests that most airway devices are easier to deploy in humans than in most simulators.

We have performed two previous evaluations of these manikins. In an evaluation of ease of LMA Classic insertion we found that the Trucorp and VBM manikins performed best, with the Ambu manikin performing worst [4]. In contrast, when examining the manikins for insertion of and ventilation thorough a wide variety of supraglottic airway devices we found that the Trucorp manikin performed best, with this and the Laerdal manikin performing considerably better than the other two overall, though this finding was not consistent for individual supraglottic airway devices [5]. The implication is that several manikins may be necessary to enable training on a wide variety of airway devices.

The methodology is useful for future evaluations of airway devices, both manikins and supraglottic airways, and has implications for interpretation of other manikin research. We conclude that although several manikins performed adequately for insertion of and ventilation through currently available laryngeal mask airways, the VBM manikin performed best overall and for several individual laryngeal masks.

Conflict of interest and financial support

This study was supported by a grant from the Difficult Airway Society, which purchased two of the manikins (VBM and Ambu). The manikin from Trucorp was donated to the study and that from Laerdal loaned for the period of the study. TMC has participated in informal evaluation of Laerdal equipment for the company. He has also received payment for lecturing from Intavent Orthofix and the LMA Company, both of which manufacture laryngeal mask airways. All laryngeal mask airways used in the study were donated by their manufacturers either free of charge or at cost price.