Critical incidents during anaesthesia are often related to airway management.1–3 A difficult airway can be predicted by the presence of certain anatomical criteria but accurate prediction remains problematical and leaves a certain rate of undetected (thus unexpected) difficult intubation.4 Unpredictable difficult laryngoscopy remains a challenge for anaesthesiologists, especially if difficult ventilation occurs. Difficult laryngoscopy occurs when the view of the laryngeal inlet is poor and corresponds to grade 2b (only arytenoids visible), 3a (only epiglottis visible and liftable), 3b (epiglottis adherent to pharynx) and 4 (no laryngeal structures seen) of the Cormack–Lehane classification: it is reported in the literature to have an incidence of 5 to 10%.1,2,5,6 Cormack–Lehane grades 2b and 3a are recognised as a ‘restricted’ laryngeal view and grades 3b and 4 as a ‘difficult’ laryngeal view.5
Accurate airway assessment should always be performed so as to provide appropriate planning and management of expected difficult intubation and to limit any unexpected difficulties, but the common clinical screening tests (Mallampati score, interincisor distance, thyrohyoid distance, chin–hyoid distance, BMI etc.) have shown low sensitivity and specificity with a limited predictive value, especially if only a single assessment method is used.7–10 Better predicting power can be obtained by combining several clinical parameters or by a single test composed of a variety of parameters, such as El-Ganzouri's test.11,12 It is also claimed that these clinical tests are often useless or difficult to perform in emergency situations and in critically ill or uncooperative patients unless dedicated tests (e.g. the Machoca score) are used.13,14
Airway assessment using ultrasound has been proposed recently as a useful, simple and noninvasive bedside tool as an adjunct to clinical methods, but to date few studies are available about the potential role of ultrasound in difficult airway evaluation and these are mostly limited to specific groups of patients (e.g. the obese).15,16 Furthermore, the data obtained do not provide definitive results and also it is not clear which specific ultrasound measurements are predictive of or related to difficult intubation.15–18
The aim of the current study was to establish whether correlations existed between ultrasound measurements of anterior cervical soft tissue thickness at two anatomical levels (thyrohyoid membrane or ‘pre-epiglottic space’ and vocal cords) and the Cormack–Lehane grade during direct laryngoscopy, and whether these measurements are useful in predicting a restricted or difficult view. We chose these two new measurements from among the various ultrasound assessments made in previous studies because of their simplicity of execution in normal clinical and in emergency settings.
The study was designed as a prospective observational single-blinded study. Eligibility criteria were adults (patients ≥18 years of age) scheduled for elective surgery, who required tracheal intubation under general anaesthesia in a teaching hospital operating theatre at Ospedali Riuniti Ancona (Ancona, Italy) from May 2017 to September 2017. Ethical approval for this study was provided by Marche's Regional Ethics Committee (CERM), Ancona, ITA (Protocol No. 2016-0405), Chairperson Prof Marcello D’Errico, on 23 March 2017. Informed written consent was obtained. The study conformed to the Declaration of Helsinki and Good Clinical Practice guidelines.
The following patients were excluded from the study: patients who required rapid sequence induction of anaesthesia, uncooperative or pregnant patients, patients with a history of previous difficult intubation, expected difficult laryngoscopy due to limited cervical spine motility, maxillofacial anomalies, interincisor distance less than 3 cm, thyromental distance less than 6 cm and other parameters as stated in The Italian Society of Anesthesia Analgesia Reanimation and Intensive Care (SIAARTI) Guidelines,3 were excluded from the study. The exclusion of patients with clinical criteria of predicted difficult intubation should permit specifical recognition of unpredicted difficult laryngoscopy using the ultrasound measurements.
Pre-operative evaluation was performed before surgery, demographic variables were collected and clinical screening tests to predict a difficult airway were performed. After adequate training (20 senior-assisted bedside scanning tests), two anaesthetists performed the ultrasound measurements at the pre-operative visit using a Caris Plus ultrasound machine (ESAOTE, Genoa, Italy) with a 10 to 13-MHz linear ultrasound transducer. We chose to perform the ultrasound examination with the patient supine and with the head in a neutral position without neck extension (Fig. 1a) for two reasons: first, it is simple to achieve in a standardised manner, and second, it could be used in emergency settings such as trauma.
The ultrasound probe was placed in the transverse plane and was used to measure the thickness of the anterior cervical soft tissues at two levels [the thyrohyoid membrane (pre-epiglottic space) and the vocal cords (laryngeal inlet)]. At the level of thyrohyoid membrane, the epiglottis is visible as a hypoechoic curvilinear structure with its posterior border demarcated by a bright hyperechoic linear air-mucosal interface. In this plane, the visualisation obtained, including the thyrohyoid muscles, provides an image we called the ‘small face sign’ (Fig. 1b). The two vocal cords with arytenoids appear as hyperechoic lateral V-shaped structures (Fig. 1c). Protrusion of the tongue or swallowing help to identify the epiglottis, whereas identification of the vocal cords is facilitated by observing their movements during breathing or phonation.
As in previous studies,15,29 ultrasound measurement of the distance from the skin was performed at each level in the median axis, 10 mm to the right and 10 mm to the left. The area connecting these points was calculated to quantify the thickness of the soft tissues better. The distance from the skin to the epiglottis (mDSE) in the median axis performed at the level of the thyrohyoid membrane was recorded, as was the median distance from the skin to the apex of the vocal cords (mVC).
General anaesthesia was administered according to a standard protocol. The ECG, SpO2, noninvasive blood pressure, capnography and accelerometric neuromuscular monitoring of train-of-four (TOF) were employed. Depth of anaesthesia was also measured with an EntropyTM monitor (GE Healthcare, Helsinki, Finland). Pre-oxygenation was performed for 3 min, propofol (2 mg kg−1) was administered and then a neuromuscular blocker (rocuronium 1 mg kg−1) was given after checking that facemask ventilation was adequate. When the anaesthetic plane was deep enough (State Entropy below 50) and neuromuscular blockade was deemed as adequate, based on TOF, conventional direct laryngoscopy with a Macintosh blade was performed to evaluate the Cormack–Lehane grade and to perform tracheal intubation.
The laryngoscopic view was graded by an anaesthetist with more than 5 years of experience, who had not performed the ultrasound assessments and who was blinded to the results of the assessments. For each patient, only the best attempt at direct laryngoscopy was considered.: obtained after optimising position, complete muscle relaxation (TOF = 0) and, if necessary, external laryngeal manipulation. A decrease of SpO2 to less than 92% was a criterion for abandoning the procedure, and a maximum of three attempts were permitted before declaring intubation failure with direct laryngoscopy. A Cormack–Lehane grade 1 or 2a was classified as ‘easy’ laryngoscopy and grades between 2b and 4 as ‘restricted or difficult’ laryngoscopy.
According to the literature,16,17 ultrasound measurements should predict difficulty in at least 80% of the intubations that are really difficult (with a Cormack–Lehane grade ≥2b). As the incidence of difficult intubation is about 5 to 10% of all intubations,1–3,6 it would be necessary to study at least 244 patients to obtain statistically significant differences between the two groups, accepting an alpha error of 0.05 and a beta error of 0.20.
Data analysis was performed using the Medcalc 126.96.36.199 statistical program (Medcalc software, Ostend, Belgium). Kolmogorov–Smirnov's test was used to test the normality of distribution. Receiver-operating characteristic curves (ROC) were used to determine the ‘difficulty prediction capability’ of each sonographic measurement and to assess the optimal cut-off scores. To allow for comparisons between ‘restricted/difficult’ airway and ‘easy’ airway groups, a two-sided Student t test and Fisher's exact test were employed as appropriate. The results were averaged, and the data are presented as mean ± SD for each variable with continuous data. Values of P less than 0.05 were considered as statistically significant.
A total of 301 patients aged between 18 and 90 years (mean 57.2 ± 17.2), 156 males and 145 females, were enrolled in the study (Fig. 2). The mean BMI was 25.8 ± 5.3 kg m−2. Direct laryngoscopy was classified as ‘easy’ in 273 (90.6%) patients: 189 (62.7%) patients with a Cormack–Lehane grade 1, and 84 (27.9%) patients with a Cormack–Lehane grade 2a.
The incidence of restricted or difficult laryngoscopy (Cormack–Lehane ≥ 2b) was 9.2% (28 patients) with a Cormack–Lehane grade 2b in 20 patients (6.6%) a Cormack–Lehane 3a in six patients (1.9%) and a Cormack–Lehane grade 3b in two patients (0.6%) (Table 1). In these cases, tracheal intubation was declared as ‘unexpected difficulty’. Tracheal intubation was achieved using direct laryngoscopy with a Macintosh blade and bougie in 17 patients, and 11 patients (3.6% of total study population and 39% of difficult laryngoscopy group) were intubated using alternative devices: four patients by videolaryngoscopy, four with a Bonfils fibrescope and one using an intubating laryngeal mask airway (ILMA Fast-Trach). Intubation failed in two patients (with a Cormack–Lehane grade of 3b); these patients were awakened and underwent successful intubation using a flexible bronchoscope. No complications occurred in any patient.
The correlation between ultrasound measurements and Cormack–Lehane grade laryngeal view (and its ability to predict easy or restricted/difficult laryngoscopy) is shown by comparison of the area under the ROC curves (AUC, Fig. 3). The best predictors of restricted/difficult laryngoscopy (Cormack–Lehane ≥ 2b) were the mDSE at the level of the thyrohyoid membrane (pre-epiglottic space) [AUC 0.906; 95% confidence interval (CI) 0.86 to 0.93] and the pre-epiglottic area (PEA) (AUC 0.93; CI 0.89 to 0.95). The cut-off values with the highest sensitivity and specificity were 2.54 cm (sensitivity 82%; specificity 91%) and 5.04 cm2 (sensitivity 85%; specificity 88%), respectively. No correlation was found between difficult laryngoscopy and ultrasound assessment at the level of the vocal cords (AUC 0.54, CI 0.48 to 0.60, sensitivity 53%, specificity 66%).
Comparing the ROC curve for the predictive power of the mDSE between male and female patients, the cut-off value of 2.54 cm showed an AUC of 0.94 (CI 0.90 to 0.97) in the female group with a sensitivity of 100% and a specificity of 83%. In male patients, the AUC was 0.86 (CI 0.80 to 0.91) with a sensitivity of 75% and specificity of 92% (Fig. 3).
Post hoc analysis revealed that 10 of 11 patients who underwent intubation with alternative devices after failure with a Macintosh blade were predicted, with ultrasound measurements of mDSE more than 2.54 cm and PEA more than 5.04 cm2 (Table 2).
We found a high correlation with the ROC curves between two ultrasound measurements (mDSE and PEA) of the anterior cervical soft tissue thickness at the level of the thyrohyoid membrane and Cormack–Lehane score at direct laryngoscopy.
In the last decade, ultrasound has become an important tool for various diagnostic or therapeutic purposes in the operating theatre and critical care settings. The use of ultrasound in airway management is quite recent. It can be used for determining the correct placement of the tracheal tube18 or the correct size of paediatric19 and double-lumen tubes,20,21 for diagnosis of upper airway obstruction and for guidance of percutaneous tracheostomy/cricothyrotomy.18 It could be used for performing laryngeal nerve blocks to facilitate awake intubation,22 for the diagnosis of postextubation stridor and for confirmation of proper laryngeal mask airway position.23,24 Results are contradictory and not uniform, but there is potential for use of ultrasound in airway management, as summarised in excellent recent reviews.25 Some authors have used ultrasound for predicting difficult intubation, but until now, there has been low agreement and little evidence about which ultrasound assessments are the best predictors. Some authors have obtained measurements of the cervical anterior soft tissues thickness at different levels.15–17
Recently, Weidong et al.26 performed sonographic measurements of mandibular condylar mobility to predict difficult laryngoscopy.26 The condylar translation prediction criterion for difficult laryngoscopy was 10 mm or less and it correlated with Cormack–Lehane grades of at least 3. It proved to be an independent predictor by multivariate logistic regression compared with others common clinical assessments.
In our observational study, we examined the correlation between ultrasound measurements performed at two anterior cervical levels and the Cormack–Lehane grade at direct laryngoscopy. We excluded patients with certain clinical criteria predictive of difficult intubation as we wished to study the ultrasound measurements as independent variables in patients in whom direct laryngoscopy was not predicted to be difficult by clinical assessment. In this way, ultrasound might help to identify the ‘clinically unexpected’ difficult airway. Ultrasound assessments were performed at the level of the thyrohyoid membrane looking for what we called the ‘small face’ sign (Fig. 1b) and of the vocal cords (Fig. 1c).
We hypothesised that increasing thickness at the level of the pre-epiglottic space could affect the ability to visualise the glottis with a Macintosh blade at direct laryngoscopy, especially when associated with a reduced distance between the skin and vocal cords. It could be explained with the anatomical model recently described by Greenland27 (Fig. 4). That model suggested that the upper airways are shaped by two curves, an oropharyngeal or ‘primary’ curve and a pharyngo-glottic-tracheal or ‘secondary’ curve. Adequate laryngoscopic visualisation requires that both curves have to be aligned with the visual axis. A large distance from skin to epiglottis could be the result of a higher upwards concavity of the primary curve leading to bad glottic visualisation. In the same way, a small distance between the skin and the vocal cords could produce a high downward concavity of the secondary curve, worsening the direct laryngoscopic view.
Our results showed that a median distance from skin to epiglottis (mDSE) above 2.5 cm and a PEA above 5.0 cm2 at the level of thyrohyoid membrane may predict a Cormack–Lehane grade at least 2b at direct laryngoscopy; this could be due to an increase in the upward concavity of the primary curve that could worsen the line of sight during direct laryngoscopy. In contrast, no correlation was found between difficult laryngoscopy and ultrasound assessments at the level of the vocal cords in relation to changes of the secondary curve's steepness. This result is in contrast with the study of Preethi et al.,28 who found that a thickness value more than 0.2 cm at the level of the vocal cords was a good predictor of difficult intubation; however, these authors did not exclude patients with anatomical characteristics predictive of difficult intubation, such as Mallampati score more than 3.
We also compared the ROC curves of the median pre-epiglottic thickness (mDSE) between male and female patients, hypothesising a possible correlation with anatomical laryngeal differences between sexes. We found that the same cut-off value of 2.5 cm, which was the best predictor of a restricted/difficult laryngoscopy, seemed to have the best performance in female patients in terms of sensitivity (100 vs. 75%), whereas in men, there was greater specificity (92 vs. 83%) (Fig. 3). The ultrasound measurements in the female patients could be considered to be a better sensitivity model to reduce the number of unpredicted difficult laryngoscopies compared with male patients. This difference could correspond to a differently shaped thyroid cartilage or to a different pattern of neck fat deposition in women.
Ezri et al.15 found that an abundance of soft tissue anterior to the vocal cords as measured by ultrasound was a good independent predictor of difficult laryngoscopy, but this study referred only to an obese patient population, and it was found to be in accordance with the predictive clinical index of an increased neck circumference. In our study sample, median BMI was 25.8 kg m−2; therefore, the results are not strictly comparable.
In another study, Komatsu et al.16 found that the anterior neck soft tissue thickness measured with ultrasound at the level of the vocal cords in a sample of 64 obese patients was not a good predictor of difficult laryngoscopy. Subsequently, in 2011, a pilot study by Adhikari et al.29 to develop a protocol of three-level ultrasound measurements in the anterior cervical region obtained encouraging but not definitive results due to the small sample size examined (50 patients). In that study, the cut-off value of 2.8 cm for anterior neck soft tissue thickness at the thyrohyoid membrane was a good indicator to predict difficult laryngoscopy, not far from our 2.54-cm value calculated on a larger patient sample.
Pinto30 showed that ultrasound measurements of anterior neck soft tissue at the level of the thyrohyoid membrane (DSE > 2.75 cm) could improve the predictive power of current standard clinical screening tests when combined with a modified Mallampati score in a simple decisional tree. On the contrary, Adhikari et al.29 demonstrated that the Mallampati clinical screening test did not correlate with ultrasound measurements.
In our study, the incidence of restricted/difficult laryngoscopy was 9.2%. It was recorded in 28 patients and intubation with a Macintosh blade failed in 3.6% (11/301 patients; Fig. 2 and Table 2). Rescue intubation was obtained with alternative devices as described above. Obtaining an adequate laryngoscopic view is essential for successful intubation using direct laryngoscopy, but difficult laryngoscopy and difficult intubation are not synonymous. Successful intubation depends also on the operator's expertise, and the patient's characteristics and circumstances. Asai et al.31 showed that of three patients with a Cormack–Lehane grade 4, only one was difficult to intubate, whereas one was easy and one moderately difficult to intubate. Of 68 patients with a Cormack–Lehane grade 3, five were difficult to intubate, 50 moderate and 13 easy. Therefore, the ‘restricted view’ (Cormack–Lehane grade 2b or 3a) at direct laryngoscopy might become a critical factor in some situations such as in an emergency, when a double-lumen tube must be inserted or when the person who is performing laryngoscopy is not an experienced anaesthetist, for example.
As shown in Fig. 5, we also tested the predictive capacity of the mDSE for the use of alternative devices after intubation failure with the Macintosh laryngoscope. As shown in the chart, patients intubated with the Macintosh blade were allocated in column zero and patients requiring alternative devices in column one. Almost all patients in column one had a mDSE at least 2.54 cm, whereas most patients in Column zero had a mDSE lower than the cut-off. The sensitivity of the test was 91.7% and the specificity was 87.9%. The ultrasound assessment could thus be considered a good predictor of difficult intubation, not only of difficult laryngoscopy, and it might indicate a need for preprocedural preparation of alternative devices.
Our data analyses indicate that ultrasound measurements at these levels can be obtained in less than 2 min, supporting their potential utility not only in the normal clinical setting but also in the critical care setting, especially when the patient cannot cooperate or should be maintained with the head in a neutral position for cervical spine protection, and therefore, the common clinical screening tests could not be performed easily. However, further studies are necessary to test the predictive capability of ultrasound for airways management in an emergency setting.
We excluded patients with a difficult intubation predicted by means of clinical tests. This might create potential selection bias, but we thought that this choice might allow us to study the ultrasound measurements as independent predictive assessments of a difficult airway and to identify difficult intubation in patients with no clinically predictable difficulty.
The next step of validation should involve patients with predictable difficult laryngoscopy or intubation, so further studies are necessary to compare ultrasound measurements with all the standard clinical screening tests of a difficult airway using multivariate statistical analysis.
We chose to study only two new ultrasound measurements and not other measures considered in previous studies, as they are very simple to obtain. Furthermore, we had undertaken a previous pilot study in a smaller sample of patients,32 in which we analysed the correlation between Cormack–Lehane grade and ultrasound measures at three levels: epiglottis (mDSE; PEA); vocal cords and hyoid bone. We found no correlation for the measures at the level of the hyoid bone, a mild correlation at the level of vocal cords and a high correlation for mDSE and PEA.
Ultrasound assessment of pre-epiglottic tissue thickness at the level of the thyrohyoid membrane may be useful to predict restricted/difficult direct laryngoscopy and difficult intubation with the Macintosh blade. Airway evaluation remains of paramount importance, which is also a clear statement and decisional pathway of SIAARTI guidelines, despite the knowledge that it might have limitations in terms of sensitivity and specificity. Developing a test which might improve overall sensitivity while maintaining a high specificity might result in a lower numbers of false negatives and a lower incidence of false positives, improving patient safety and allowing the best use of resources.
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: none.
Conflicts of interest: none.
1. Peterson GN, Domino KB, Caplan RA, et al. Management of the difficult airway. A closed claims analysis. Anesthesiology
2. Cook TM, Woodall NC, 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
3. Petrini F, Accorsi A, Adrario E, et al. S.I.A.A.R.T.I. Guidelines. Recommendations for airway control and difficult airway management. Minerva Anestesiol
4. Pandit JJ, Heidegger T. Putting the ‘point’ back into the ritual: a binary approach to difficult airway prediction. Anaesthesia
5. Cook TM. A new practical classification of laryngeal view. Anaesthesia
6. Frerk VS, Mitchell AF, McNarry C, et al. Difficult Airway Society 2015 guidelines for management of unanticipated difficult intubation in adults. Br J Anaesth
7. Vannucci A, Cavallone LF. Bedside predictors of difficult intubation: a systematic review. Minerva Anestesiol
8. Naguib M, Scamman FL, O'Sullivan C, et al. Predictive performance of three multivariate difficult tracheal intubation models: a double-blind, case controller study. Anesth Analg
9. Langeron O, Cuvillon P, Ibanez-Esteve C, et al. Prediction of difficult tracheal intubation: time for a paradigm change. Anesthesiology
10. Frova G, Sorbello M. Algorithms for difficult airway management: a review. Minerva Anestesiol
11. El-Ganzouri AR, McCarthy RJ, Tuman KJ, et al. Preoperative airway assessment: predictive value of a multivariate risk index. Anesth Analg
12. Kheterpal S, Healy D, Aziz MF, et al. Incidence, predictors, and outcome of difficult mask ventilation combined with difficult laryngoscopy: a report from the Multicenter Perioperative Outcomes Group. Anesthesiology
13. De Jong A, Molinari N, Terzi N, et al. Early identification of patients at risk for difficult intubation in the intensive care unit: development and validation of the MACOCHA score in a multicenter cohort study. Am J Respir Crit Care Med
14. Reed MJ, Dunn MJ, McKeown DW. Can an airway assessment score predict difficulty at intubation in the emergency department? Emerg Med J
15. Ezri T, Gewurtz G, Sessler DI, et al. Prediction of difficult laryngoscopy in obese patients by ultrasound quantification of anterior neck soft tissue. Anaesthesia
16. Komatsu R, Sengupta P, Wadhwa A, et al. Ultrasound quantification of anterior soft tissue thickness fails to predict difficult laryngoscopy in obese patients. Anaesth Intensive Care
17. Wojtczak JA. Submandibular sonography: assessment of hyomental distances and ratio, tongue size, and floor of the mouth musculature using portable sonography. J Ultrasound Med
18. Kundra P, Mishra SK, Ramesh A. Ultrasound of the airway. Indian J Anaesth
19. Shibasaki M, Nakajima Y, Ishii S, et al. Prediction of pediatric endotracheal tube size by ultrasonography. Anesthesiology
20. Sustic A, Miletic D, Protic A, et al. Can ultrasound be useful for predicting the size of a left double-lumen bronchial tube? Tracheal width as measured by ultrasonography versus computed tomography. J Clin Anesth
21. Merli G, Guarino A, Della Rocca G, et al. SIAARTI Studying Group on Difficult Airway. Recommendations for airway control and difficult airway management in thoracic anesthesia and lung separation procedures. Minerva Anestesiol
22. Chin-Hung L, Wei-Chun C, Yao-lin Y. A new method for ultrasound-guided laryngeal superior nerve block. Tzu Chi Medical Journal
23. Ding LW, Wang HC, Wu HD, et al. Laryngeal ultrasound: a useful method in predicting postextubation stridor. A pilot study. Eur Respir J
24. Wojtczak JA, Cattano D. Laryngo-tracheal ultrasonography to confirm correct endotracheal tube and laryngeal mask airway placement. J Ultrason
25. Kristensen MS. Ultrasonography in the management of the airway. Acta Anaesthesiol Scand
26. Weidong Y, Yumei Z, Bin W, et al. Can mandibular condylar mobility sonographic measurements predict difficult laryngoscopy? Anesth Analg
27. Greenland KB. Airway assessment based on a three column model of direct laryngoscopy. Anaesth Intensive Care
28. Preethi BR, Pankaj P, Kolli SC. Ultrasonography: a viable tool for airway assessment. Indian J Anaesth
29. Adhikari S, Zeger W, Schmier C, et al. Pilot study to determine the utility of point of-care ultrasound in the assessment of difficult laryngoscopy. Acad Emerg Med
30. Pinto J. Predicting difficult laryngoscopy using ultrasound measurement of distance from skin to epiglottis. J Crit Care
31. Asai T, Koga K, Vaughan RS. Respiratory complications associated with tracheal intubation and extubation. Br J Anaesth
32. Falcetta S, Cavallo S, Pelaia P. Ultrasound measurements as predictors of difficult direct laryngoscopy. Trends Anaesthesia Crit Care