Secondary Logo

Journal Logo

Original Article

Prediction of difficult tracheal intubation

Iohom, G.; Ronayne, M.; Cunningham, A. J.

Author Information
European Journal of Anaesthesiology (EJA): January 2003 - Volume 20 - Issue 1 - p 31-36
  • Free


The incidence of difficult laryngoscopy or intubation varies from 1 to 18%, and failed intubation has been identified as an anaesthesia-related cause of death or permanent brain damage [1]. Preoperative screening to identify patients in whom tracheal intubation may be difficult or impossible may reduce mortality and morbidity [2]. Preoperative bedside screening tests including the Mallampati oropharyngeal classification (MP), thyromental (TM) and sternomental (SM) distances, have been shown to be neither sensitive nor specific enough for routine clinical use [2,3]. Conflicting data have been reported about the efficacy of combining two or more of these tests and their ability to predict difficult intubation [4-6]. The objective of the present study was to investigate if a combination of the Mallampati classification of the oropharyngeal view with either the thyromental or the sternomental distance measurement would improve the ability to predict difficult intubation compared with each test alone.


The prospective single-blinded study was performed at Beaumont Hospital, Dublin, Ireland, and was approved by the institutional Ethics Committee. Written informed consent was obtained from 212 patients aged >18 yr scheduled for elective surgical procedures requiring tracheal intubation. Patients with known airway abnormalities or with obvious head and neck pathologies were excluded. The airway assessment was performed by the primary investigator in each case to avoid interobserver variability.

Preoperatively, the size of the tongue relative to the oral cavity was assessed according to the Mallampati criteria with slight modifications to avoid ambiguity as described by Tse and colleagues [6]. The test was performed using a pen torch with the patient in the sitting position, mouth wide open with the tongue protruding and the patient not phonating. Class I indicates that the faucial pillars, soft palate and uvula can be seen; Class II indicates that the faucial pillars are not visible but the uvula remains partly visible; Class III indicates that the uvula is completely masked by the base of the tongue and that the posterior pharyngeal wall is not visible; Class IV indicates that only the hard palate is visible.

The thyromental distance was measured as the straight-line distance (cm) from the thyroid notch to the lower border of the mentum with the head fully extended on the neck and the mouth closed. A ruler was used and distances were approximated to the nearest 0.5 cm (Fig. 1)[7]. The sternomental distance was measured as the distance (cm) between the upper border of the manubrium sterni and the bony point of the chin with the head in full extension and the mouth closed (Fig. 2). An oropharyngeal Mallampati Class III or IV, a thyromental distance <6.5 cm and a sternomental distance <12.5 cm were selected as predictors of difficult intubation [8].

Figure 1
Figure 1:
Standard measurement of the thyromental distance.
Figure 2
Figure 2:
Standard measurement of the sternomental distance.

Before induction of anaesthesia, intravenous access and standard monitoring was established with a pulse oximeter, electrocardiogram and non-invasive arterial pressure monitoring (Datex AS/3®; Dale Corp, Madison, WI, USA). Induction of anaesthesia was performed in a conventional way with the administration of a neuromuscular blocking agent (using a technique appropriate for the individual patient and clinical circumstances and a nerve stimulator to assess the adequacy of neuromuscular blockade). Then laryngoscopy was performed by a post-fellowship anaesthetist blinded to our airway assessment using a Macintosh blade and the view was classified according to the method described by Cormack and Lehane [9]. Thus, Grade 1 indicates a full view of the glottis; Grade 2, a partly exposed glottis with anterior commissure not seen; Grade 3, only the epiglottis is seen; Grade 4, that the epiglottis is not seen. An intubation was considered difficult if the patient had a Cormack and Lehane laryngoscopy Grade of 3 or 4, or if a bougie was required to aid intubation in patients with a laryngoscopy Grade of 2.

The preoperative assessment data and intubation findings were correlated to determine the accuracy of the three tests in predicting difficult intubation. Each test, alone and together with various combinations, was evaluated by calculating the sensitivity (the proportion of difficult tracheal intubations correctly predicted to be difficult), specificity (the proportion of easy tracheal intubations correctly predicted to be easy), positive predictive value (the proportion of predicted difficult intubations which actually proved to be difficult) and negative predictive value (the proportion of predicted easy intubations which actually proved to be easy) using the following formulae: EQUATION

where TP is the true-positive, TN is the true-negative, FP is the false-positive and FN is the false-negative.


Table 1 shows the gender distribution within the sample as well as in the difficult intubation group. Altogether, 51% of the 212 patients were males with 49% females. Of the total intubations, 9% were difficult as defined by the criteria mentioned above. A significantly higher proportion of difficult tracheal intubations was found in males. There was no instance of failed intubation.

Table 1
Table 1:
Gender characteristics of the sample.

Figure 3 shows the frequency distribution of the preoperative oropharyngeal grading within the sample, continuous areas representing the proportion of difficult intubations. Only five of 96 patients belonging to the Mallampati Class I group proved to be difficult, whereas two of three tracheal intubations in the Mallampati Class IV group were difficult.

Figure 3
Figure 3:
The top of each column represents the actual difficult tracheal intubations within each Mallampati class. MP: Mallampati.

The frequency distribution of the thyromental distances within the sample (Fig. 4) shows that the majority of patients had a thyromental distance of 8 cm. The incidence of difficult intubation increased significantly with the decrease of the thyromental distance. Of note, none of the patients with a thyromental >9 cm had a difficult intubation.

Figure 4
Figure 4:
The top of each column represents the actual difficult tracheal intubations within each group of thyromental distances.

The frequency distribution of the sternomental distances within the study population (Fig. 5) confirms that all three patients with a sternomental <11 cm had a difficult intubation, whereas there was no difficulty in intubating patients with a sternomental >17.5 cm.

Figure 5
Figure 5:
The top of each column represents the actual difficult intubations within each group of sternomental distances.

Figure 6 shows the combined preoperative data and incidence of difficult intubation. All six patients with a Mallampati Class III/IV who had thyromental <6.5 cm and thyromental <12.5 cm proved to be difficult, whereas only 14 of 192 patients with Mallampati Class I/II and thyromental >6.5 cm and sternomental >12.5 cm were difficult.

Figure 6
Figure 6:
Incidence of difficult tracheal intubation in the two groups: (a) Mallampati III/IV, TM < 6.5 cm, SM < 12.5 cm; (b) Mallampati I/II, TM > 6.5 cm, SM > 12.5 cm. MP: Mallampati classification of the oropharyngeal view; TM: thyromental distance; SM: sternomental distance. ▪: Difficult intubation; □: easy intubation.

Table 2 features the sensitivity, specificity, and positive and negative predictive values of each preoperative test alone and in various combinations. Combining Mallampati with either thyromental and/or sternomental distances reduces the sensitivity but increases the specificity and positive predictive value to 100%. Combining thyromental with sternomental increases the positive predictive value to only 67%.

Table 2
Table 2:
Data interpretation.


Management of the unexpectedly difficult intubation is a challenge for every anaesthetist. Difficulty in visualizing the glottis during laryngoscopy under general anaesthesia may cause difficulty or the failure of tracheal intubation. The definition, and hence incidence of difficult intubation, varies widely in the literature. For example, Langenstein and Cunitz call an intubation difficult 'if a normally trained anaesthesiologist needs more than three attempts or more than 10 min for a successful tracheal intubation' [10]. Cormack and Lehane Grades 3 or 4 are generally regarded as 'difficult' [11,12]. The addition of the need to use a gum elastic bougie in this definition has been criticized as the threshold for use of a bougie is variable [11,13]. However, the use of a bougie was included as the anatomical factors, which make these patients 'relatively difficult', are the same factors that in more extreme cases cause Grade 3 or 4 difficulty [8]. The incidence of difficult intubation depends on the degree of difficulty encountered showing a range of 1-18% of all tracheal intubations with about 2/10 000 to 1/1 000 000 for 'cannot ventilate-cannot intubate' situations [10].

Problems in airway management can be predicted based on previous anaesthesia records, the medical history of the patient and the physical examination of multiple anatomical features of the head and neck [1]. Several radiological measurements have been reported, but they are impractical for population screening [10,14]. The majority of anaesthetists rely on predicting difficult intubation mainly as a result of the impression gained from the end of the bed [5,15]. Several preoperative bedside screening tests have been described to predict difficult intubation such as the oropharyngeal (Mallampati) classification [16], measurement of the thyromental [17] and measurement of the sternomental distances [1,8]. Each individual test when used alone has been shown to have a low sensitivity, a low positive predictive value and a low to moderate specificity [5,6,8,11,18]. However, one of the greatest criticisms of the Mallampati test has been the problem of interobserver variation [5]. To overcome this limitation, the same investigator performed the preoperative assessment in our study under standardized conditions.

Whether combining two or more of these preoperative screening tests will improve the prediction of difficult intubation is a controversial topic in the literature. Frerk found that combining the Mallampati classification with the thyromental distance improved the specificity while maintaining the same level of sensitivity [5]. However, Tse and colleagues found a significant reduction in the sensitivity when combining the two tests, while the specificity was increased [6].

Our data suggest that the Mallampati classification used alone has a poor sensitivity (40%), poor specificity (89%) and poor positive predictive value (27%). Combining the Mallampati with either the thyromental or sternomental distances decreases sensitivity to 25 and 20% respectively, while it increases specificity and the positive predictive value to 100%. Of note, there is no significant change in the negative predictive value (93%). The combination of the thyromental and sternomental distances alone has a sensitivity of 33%, while the positive predictive value is increased to only 67%.

It is well known that mobility and space are the two prerequisites for a successful direct laryngoscopy - we need mobility of the atlanto-occipital joint to extend the head, that of the temporomandibular joint to open and subluxate the jaw; and we also need sufficient oropharyngeal space to displace the tongue [19]. A positive Mallampati test indicates the possible existence of a limited mouth opening or a limited oropharyngeal space. However, it does not include the assessment of the atlanto-occiptal joint mobility. The sternomental distance may be a good indication of maximum head extension, enabling a more accurate assessment of head extension than subjective assessment and avoiding the need for radiological examination. The thyromental distance also gives a measure of head extension in addition to the degree of recedence of the jaw (retrognathia). Therefore, combining the Mallampati test with either of sternomental or thyromental distances it is likely to improve the overall predictability of difficult intubation.

A serious limitation to this study is that several different laryngoscopists undertook laryngoscopy and it is possible that a uniform intubation position was not achieved. However, all staff participating in the study had a minimum of 4 yr' experience at senior registrar/consultant level. It is also known that unexpected problems with the epiglottis and glottic inlet are potential sources of danger that are difficult to predict preoperatively. This is one of the evident reasons why no screening test is 100% sensitive and 100% specific. Familiarity with a difficult intubation drill and access to a difficult intubation trolley (range of laryngoscopes, endotracheal tubes, laryngeal masks and a fibreoptic bronchoscope) should be readily available at all times.

In conclusion, the findings of our study suggest that the Mallampati classification in conjunction with thyromental and sternomental distances may be a useful routine screening test for preoperative prediction of difficult intubation. Clearly, no simple test can be used alone. These three simple bedside tests can be easily performed at the routine preoperative visit; any patient identified as having both a Class III or IV view of the oropharynx and a thyromental distance <6.5 cm or a sternomental distance <12.5 cm can be expected to present difficulty with intubation of the trachea. In this group of patients, advanced planning of the anaesthetic is mandatory so that the presence of experienced personnel can be guaranteed. If local anaesthetic techniques are unsuitable or if a general anaesthetic is essential, the use of an awake intubation technique may be indicated.


The authors thank all the consultants and senior registrars who assisted in collecting the data, and Ms Denise O'Mara for the preparation of the images.


1. Randell T. Prediction of difficult intubation. Acta Anaesthesiol Scand 1996; 42: 136-137.
2. Voyagis GS, Kyriakis KP, Dimitriou V, Vrettou I. Value of oropharyngeal Mallampati classification in predicting difficult laryngoscopy among obese patients. Eur J Anaesthesiol 1998; 15: 330-334.
3. Bergler W, Maleck W, Baker-Schreyer A, Ungemach J, Petroianu G, Hormann K. The Mallampati score. Prediction of difficult intubation in otolaryngologic laser surgery by Mallampati score. Anaesthesist 1997; 46: 437-440.
4. Koay CK. Difficult tracheal intubation-analysis and management in 37 cases. Singapore Med J 1998; 39: 112-114.
5. Frerk CM. Predicting difficult intubation. Anaesthesia 1991; 46: 1005-1008.
6. Tse JC, Rimm EB, Hussain A. Predicting difficult endotracheal intubation in surgical patients scheduled for general anaesthesia: a prospective blind study. Anesth Analg 1995; 81: 254-258.
7. Estebe JP. Prediction of difficult intubation: are we talking about the same thing? Can J Anesth 2001; 48: 719-720.
8. Savva D. Prediction of difficult intubation. Br J Anaesth 1994; 73: 149-153.
9. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia 1984; 39: 1105-1111.
10. Langenstein H, Cunitz G. Difficult intubation in adults. Anaesthesist 1996; 45: 372-383.
11. Ramadhani SAL, Mohamed LA, Rocke DA, Gouws E. Sternomental distance as a sole predictor of difficult laryngoscopy in obstetric anaesthesia. Br J Anaesth 1996; 77: 312-316.
12. Williams KN, Carli F, Cormack RS. Unexpected difficult laryngoscopy: a prospective survey in routine general surgery. Br J Anaesth 1991; 66: 38-44.
13. Higashizawa T, Bito H, Nishiyama T, Sakai T, Konishi A. Assessment of the practice of endotracheal intubation by levering laryngoscope in teaching of undergraduate medical students. Masui 1997; 46: 1255-1258.
14. Naguib M, Malabarey T, AlSatli RA, Al Damegh S, Samarkandi AH. Predictive models for difficult laryngoscopy and intubation. A clinical, radiologic and three-dimensional computer imaging study. Can J Anesth 1999; 46: 748-759.
15. Duchynski R, Brauer K, Hutton K, Jones S, Rosen P. The Quick Look Airway Classification: a useful tool in predicting the difficult out-of-hospital intubation: experience in air medical transport program. Air Med J 1998; 17: 46-50.
16. 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.
17. Patil VU, Stehling LC, Zaunder HL. Fibreoptic Endoscopy in Anaesthesia. Chicago, USA: Year Book, 1983.
18. Butler PJ, Dhara SS. Prediction of difficult laryngoscopy: an assessment of the thyromental distance and Mallampati predictive tests. Anaesth Inten Care 1992; 20: 139-142.
19. Chou H-C, Wu T-L. Thyromental distance - shouldn't we redefine its role in the prediction of difficult laryngoscopy? Acta Anaesthesiol Scand 1998; 42: 136-137.


© 2003 European Academy of Anaesthesiology