Secondary Logo

Journal Logo

Airway management

A survey of anaesthetic practice in predicting difficult intubation in UK and Europe

McPherson, Duncan; Vaughan, Ralph S.; Wilkes, Antony R.; Mapleson, William W.; Hodzovic, Iljaz

Author Information
European Journal of Anaesthesiology: May 2012 - Volume 29 - Issue 5 - p 218-222
doi: 10.1097/EJA.0b013e32835103e6


Tracheal intubation is difficult in 1–4% of patients with a seemingly normal airway.1 Consequently, unexpected difficulty in tracheal intubation is an intermittent and often terrifying problem for all practising anaesthetists. Great efforts have been made in describing preoperative assessment tests to predict a difficult laryngeal view or difficult tracheal intubation.2–7 However, many of these have methodological problems or reveal low sensitivities and specificities. Yentis8 has explained how achieving a high positive predictive value is very difficult, partly because of the rarity of a difficult intubation. The author suggested that using any of the available tests is a pointless exercise in terms of actual prediction, but that the usefulness arises from generally focusing attention on the airway. We could find no published evidence of how frequently these predictive tests are used or how useful they are perceived to be by anaesthetists. We, therefore, decided to ask the UK group and non-UK anaesthetists attending Euroanaesthesia 2006 about their current practice in predicting difficult tracheal intubation.


Twelve airway tests were identified6 and compiled into a questionnaire, using the r4 software package (Formic Ltd, Kingston, UK). We asked how often each airway test was used and how useful each test was thought to be. We also asked about features of the patient's history and how important these were to anaesthetists. In addition, we asked if there were differences of approach to airway assessment before general and regional anaesthesia.

The questionnaire was distributed by hand to delegates attending the Annual Scientific Meeting of the European Society of Anaesthesiology in Madrid, Spain in 2006 – Euroanaesthesia 2006. The questionnaires were collected at the end of each day. All selected anaesthetists were practising anaesthetists, at any grade. A random selection of UK anaesthetists was drawn up using the Medical Directory 2005/06 (Informa Healthcare, London, UK). All selected anaesthetists were practising at the consultant grade. We used the SPSS version 14 (SPSS Inc., Chicago, Illinois, USA) to select 700 anaesthetists randomly. A questionnaire was mailed to each member in this sample. A second questionnaire was sent if they had not replied within 3 months. Each questionnaire was coded and contained no personal details. Thirty-two questionnaires from the UK survey were excluded from the analysis because they were completed by anaesthetists who no longer practised clinical anaesthesia. The remainder formed the ‘UK group’ of anaesthetists. A corresponding ‘European Union (EU) group’ comprised all the responses collected at the European meeting except those from UK anaesthetists. This provided a ‘clean’ comparison between UK and ‘other’ (mainly European) anaesthetists.

Statistical analysis

The results for frequency and usefulness of the different tests are ranked on scales of 1 to 5. Therefore, the appropriate formal analysis is in terms of nonparametric tests reporting summary statistics as medians and quartiles. However, with only five possible scores, a change of just one in the score of one assessment can occasionally change a median from, for instance, 1 to 2 or 4 to 3.

If the ranks are regarded simply as a set of numbers, the mean and SD can be calculated for each test and, although the individual ranks are not normally distributed, the means of the ranks assigned by over 300 anaesthetists will be very near to normally distributed.9 Indeed, one recommended method10 of generating a normal distribution is to calculate the means of repeated groups of just 12 randomly selected zeroes and ones.


Response rate and experience

The response rate from the UK group of anaesthetists was 69% (481 of 700). The response rate from the EU group was 77% (407 of 530). The years of anaesthetic experience {median [interquartile range (range)]} of the UK and EU groups were 24 [19–29 (9–43)] and 15 [8.5–22 (1–44)] years, respectively.


Both UK and EU anaesthetists used the mouth opening and Mallampati tests much more frequently than any other test (Table 1). The subluxation, thyromental distance and Dalalkin tests all came in the first five of frequency but in a slightly different order between the UK and EU groups of anaesthetists. The remaining seven tests were mostly used much less frequently. UK anaesthetists used most of the tests less frequently than their European colleagues.

Table 1
Table 1:
Frequency of use of the individual airway assessment tests by the UK and European Union (EU) groups of anaesthetists in which the ‘EU group’ comprised non-UK anaesthetists who attended Euroanaesthesia 2006


There was much less variation between tests in terms of perceived usefulness (Table 2) (range of means less than 2 : 1) than in terms of frequency of use (nearly 4 : 1). No single test was identified as being especially useful. Both groups of anaesthetists agreed that the Nodding Donkey test is clearly less useful than any other. The UK anaesthetists found most of the tests less useful than their European colleagues (Table 2).

Table 2
Table 2:
Usefulness of the airway assessment tests as considered by the UK and European Union (EU) groups of anaesthetists

Differences in response between general and regional anaesthesia

For general anaesthesia, almost all anaesthetists would always ask about previous general anaesthesia (Table 3). However, they were less likely to assess the airway (by one or more of the 12 tests): fewer responses of ‘always’, more of ‘sometimes’ or ‘occasionally’. The same was true for airway assessment before regional anaesthesia unless it was for lower segment caesarean section (LSCS) (Table 3). This pattern was common to the UK and EU groups of anaesthetists.

Table 3
Table 3:
Frequency of use of the airway assessment tests as declared by the UK group of anaesthetists and the European Union (EU) group of anaesthetists


Devising an appropriate plan of anaesthetic management is an essential part of the patient's assessment before anaesthesia.11 A report of the recently published fourth National Audit Project of the Royal College of Anaesthetists (NAP4) states that ‘failure to assess the airway is a failure in professional duty’.12 The Difficult Airway Society (DAS) guidelines13 say that an appropriate clinical examination should include assessment of the airway.

Our findings suggest that clinicians do not always perform preoperative airway assessment tests. Thirty-three percent of the EU group and 44% of the UK group of anaesthetists failed always to assess the airway before general anaesthesia (Table 3). Furthermore, 52% of the EU group and 62% of the UK group of anaesthetists failed always to assess the airway before regional anaesthesia (Table 3). Frequency of use of individual tests varied widely (Table 1) and clinicians appear to be uncertain about their usefulness (Table 2). No matter what type of anaesthesia is involved, it is difficult to provide an explanation for the lack of airway examination. Is it that experienced anaesthetists are confident that they can manage all airways? Is this confidence misplaced? NAP412,14 evaluated the factors that contributed to major adverse events: death, permanent disability such as brain damage, unplanned surgical airway and unanticipated admission to an ICU following difficulty in airway management. Particularly, poor airway assessment contributed significantly to poor airway outcomes.15 The airway assessment was not recorded in a quarter to a third of cases with major adverse airway events which happened during anaesthesia. The audit also found that, when the tests were performed beforehand, the airway difficulty was anticipated correctly in the majority of cases. Although our survey was completed in 2007, the findings of NAP4 suggest that very little has changed in the attitude of anaesthetists towards airway assessment tests.

The results for the frequency of ‘always’ assessing the airway before regional anaesthesia are concerning (Table 3). If a regional technique fails to supply sufficient analgesia, the alternatives are either to abandon the planned procedure or convert urgent procedures to general anaesthesia. Furthermore, if the patient is to receive regional anaesthesia for a LSCS then, given the accepted alteration in airway characteristics at term, failing to examine the airway may have serious consequences if general anaesthesia becomes necessary. A recent article16 examining the changes in Mallampati class during pregnancy, labour and post-delivery found that ‘the incidence of Mallampati classes 3 and 4 increases during labour compared with the prelabour period and these changes are not fully reversed by 48 h after delivery’. This work confirms the absolute necessity of examining the airway before anaesthetic management in obstetric patients. Forewarned is forearmed.

In terms of individual tests, it is clear that many of them are used seldom (mean scores of 1–2 in Table 1), whereas mouth opening and Mallampati are by far the most frequent tests for both the UK and EU anaesthetist. Similarly, NAP4 reported Mallampati, mouth opening and neck mobility to be most commonly performed bedside airway assessment tests.12 It is disappointing that UK anaesthetists are less likely to use almost all the tests when compared with EU colleagues (mean UK–EU difference of score = −0.40), but perhaps this is because the UK practitioners find them less useful (Table 2, mean UK–EU difference = −0.23). The low frequency of use of the fibrescope (mean scores 1.36 and 1.69 for UK and EU anaesthetists, respectively; Table 1) is also disappointing when considering its perceived advantages. A recent study17 has shown that documented history of previous airway management difficulties is a strong predictor of subsequent airway problems. Therefore, it is reassuring that during preoperative assessment almost all anaesthetists asked about previous general anaesthesia (Table 3).

If an anaesthetist is responsible for the airway, what are the disadvantages associated with a conventional airway examination? The time needed to perform commonly used airway assessment tests (Table 1) is usually less than a minute. It is also prudent to note that airway assessment produces the same result with a patient sitting or in the supine position.18 Certain features may alert the anaesthetist to possible problems and lead to a change in management, for example, an awake intubation. Although an awake intubation may be considered unnecessary due to poor positive predictive value of the airway assessment tests, the record of the subsequent laryngoscopic examination of the secured airway is likely to make future airway management of that patient safer.

Recent meta-analyses of bedside airway assessment test performance19 found Mallampati classification, thyromental distance, sternomental distance, mouth opening and Wilson risk score to have poor-to-moderate sensitivity (20–62%) and moderate-to-fair specificity (82–97%). The meta-analyses also detected that the combination of Mallampati and thyromental distance was most useful. This is moderately different from the tests that were selected by the anaesthetists responding to our survey. Although these tests do not accurately predict airway difficulties, they nevertheless encourage anaesthetists to consider the possibility of some difficulty with subsequent airway management.

Can these apparent variations in anaesthetic practice be explained in another way? The lack of clinical assessment of the airway may reflect the belief that, because the positive predictive value is so low (below 30%), so is the usefulness.8 We could extend the example by Yentis with a Bayesian analysis.20 Bayesian analysis relies on the assumption that the estimate of the probability of an event (difficult intubation) depends on the information obtained before the event. Each piece of information (an airway test) alters the probability depending on its ability to predict (odds ratio) and whether the test is positive or negative. Thus, an airway assessment could be thought of as a series of dichotomous tests, each test increasing or decreasing the estimate of probability. Some tests are very good at predicting (high odds ratio) and these will have a large effect on the estimate (post-test probability). The converse is also true. Crucially, we must be able to give an initial estimate of probability (pre-test probability) for the test to work. The result of the first test becomes the input for the second and so on. In the editorial, Yentis8 exemplifies how a positive result with a high odds ratio still only gives a moderate risk of difficult intubation (30–50%) because, in Bayesian vocabulary, the pre-test probability is low. However, compared with the baseline estimate of 1–4% of difficulties in tracheal intubation in patients with an apparently normal airway,1 there is a considerable difference. Further, if all those tests were negative, the probability falls further, to less than 1%, producing around a 50-fold difference between a series of positive and negative tests. In standard vocabulary, the positive predictive value and negative predictive value are very different. It would be prudent to make some preparations for difficult intubation if there is a one in three chance of it happening. A less than one in a 100 chance may not require such preparations. These are only hypothetical examples, although we think them to be probably correct in their orders of magnitude. The optimal combination of tests and their exact odds ratios are not known. The UK and EU anaesthetists seem to think that using two or three tests from a selection of five or so is clinically the best combination. We believe that airway assessment should be a mandatory part of the preoperative visit. This is fully endorsed by NAP4, whose reviewers felt that airway management was poor in three-quarters of anaesthesia events and in more than 80% of deaths.21

What possible strategies can be pursued in the future? First, preoperative airway assessment tests should be performed in all patients. Second, we may have to accept that prediction currently is not very good and prepare more for the unexpected difficult airway. Finally, we should continue searching for better predictive tests, or combinations, to improve our estimates further and develop a strategy to integrate this information into clinical practice. We believe that all these can and should be pursued concurrently.


There was no assistance with the study. This study received no financial support. This work was sponsored by the Department of Anaesthetics, Intensive Care and Pain Medicine, University Hospital of Wales, Cardiff University, Cardiff, UK.

There are no conflicts of interest.


1. Benumof JL. Management of the difficult adult airway. With special emphasis on awake tracheal intubation. Anesthesiology 1991; 75:1087–1110.
2. Arne J, Descoins P, Fusciardi J, et al. Preoperative assessment for difficult intubation in general and ENT surgery: predictive value of a clinical multivariate risk index. Br J Anaesth 1998; 80:140–146.
3. Karkouti K, Rose DK, Wigglesworth D, Cohen MM. Predicting difficult intubation: a multivariable analysis. Can J Anaesth 2000; 47:730–739.
4. Mallampati SR, Gatt SP, Gugino LD, et al. A clinical sign to predict difficult tracheal intubation: a prospective study. Can Anaesth Soc J 1985; 32:429–434.
5. Savva D. Prediction of difficult tracheal intubation. Br J Anaesth 1994; 73:149–153.
6. Vaughan RS. Predicting a difficult intubation. In: Latto I, Vaughan RS, editors. Difficulties in tracheal intubation. 2nd ed. London: Elsevier; 1996. pp. 79–87.
7. Wilson ME, Spiegelhalter D, Robertson JA, Lesser P. Predicting difficult intubation. Br J Anaesth 1988; 61:211–216.
8. Yentis SM. Predicting difficult intubation: worthwhile exercise or pointless ritual? Anaesthesia 2002; 57:105–109.
9. Davies OL, Goldsmith PL. Statistical methods in research and production. 4th ed. London: Longman Group; 1977. pp. 21–23.
10. Shreider YA. The Monte Carlo method 1st series of monographs in pure and applied maths. Volume 87. London: Pergamon; 1966. pp. 17–23.
11. The Good Anaesthetist. Standards of practice for career grade anaesthetists. RCoA London [online]. 2010. [Accessed June 2011]
12. Cook T, Woodall N, Frerk C. Major complications of airway management in the United Kingdom. In: 4th National Audit Project of the Royal College of Anaesthetists and the Difficult Airway Society. The Royal College of Anaesthetists, 2011: 135–142.
13. Henderson JJ, Popat MT, Latto IP, Pearce AC. Difficult Airway Society guidelines for management of the unanticipated difficult intubation. Anaesthesia 2004; 59:675–694.
14. Woodall NM, Cook TM. National consensus of airway management techniques used for anaesthesia in the UK: first phase of the Fourth National Audit Project at the Royal College of Anaesthetists. Br J Anaesth 2011; 106:266–271.
15. O'Sullivan E, Laffy J, Pundit JJ. A rude awakening after our fourth ‘NAP’: lessons for airway management. Anaesthesia 2011; 66:331–340.
16. Boutonnet M, Faitot V, Catz A, et al. Mallampati class changes during pregnancy, labour and after delivery: can these be predicted? Br J Anaesth 2010; 104:67–70.
17. Lundstrome LH, Moller AM, Rosenstock C, et al. A documented previous difficult intubation as a prognostic test for a subsequent difficult tracheal intubation in adults. Anaesthesia 2009; 64:1081–1088.
18. Tham EJ, Guildersleve CD, Sanders LD, et al. Effects of posture, phonation and observer on Mallampati classification. Br J Anaesth 1992; 68:32–38.
19. Shiga T, Wajima Z, Inoue T, Sakamoto A. Predicting difficult intubation in apparently normal patients. Anesthesiology 2005; 103:429–437.
20. Thomas Bayes. An essay towards solving a problem in the doctrine of chances. Phil Trans. Royal Society London. 1763.
21. Cook TM, Woodall N, Frerk C. Major complications of airway management in the UK: results of the Fourth National Audit Project of the Royal College of Anaesthetists and the Difficult Airway Society. Part 1. Anaesthesia. Br J Anaesth 2011; 106:617–631.

airway; difficult airway; predicting

© 2012 European Society of Anaesthesiology