Inadequate ventilation, esophageal intubation, and difficult tracheal intubation are the most common mechanisms of respiratory-related adverse outcomes in anesthesia practice (1–5). The incidences of difficult laryngoscopy (DL), difficult intubation (DI), and difficult mask ventilation (DMV) are not well defined and are subject to physician variability, but they occur in 1.5%–13% (6–9), 1.2%–3.8% (3,10–16), and 0.01%–0.5% (8,9,17,18) of patients, respectively.
Preoperative evaluation is important in the detection of patients at risk for difficult airway management by noting anatomical landmarks and clinical factors associated with a difficult airway (2,3,5–7,9,19), but it is still questioned whether true prediction is possible (7,20–24) and which variables should be chosen for evaluation (19).
Although several studies have predicted a difficult airway with the use of single risk factors (13,25), risk factors used in combination (14,16), or multivariate analysis (6,18), our hypothesis considered that clinical and anatomical indexes lack sensitivity, positive predictive value (PPV), and accuracy, thus failing in the screening for and prediction of a difficult airway.
The aim of this study was to evaluate 1) the incidence of difficult airway in a surgical population; 2) the sensitivity and PPV of Mallampati and other clinical screening tests; 3) the correlation of those risk factors with ease of ventilation and laryngoscopic view; and, finally, 4) the correlation of the Cormack-Lehane (C-L) intubation scale score with a subjective operator intubation difficulty scale score (OIDS), to define whether difficult airway prediction is possible and to clarify whether the C-L scale is a valid system of scoring the laryngoscopic view during endotracheal intubation.
After obtaining institutional Human Investigation Committee approval, we prospectively collected data on 1956 consecutive adult patients scheduled to receive general anesthesia requiring endotracheal intubation for elective abdominal, vascular, urologic, and endocrinologic surgery. Patients aged <18 yr and those requiring a rapid-sequence induction or an awake intubation were excluded from the study. Patients gave their consent to be evaluated for difficult airway prediction and for their data to be used in this study (Fig. 1).
Before surgery, each patient received a clinical physical evaluation with an emphasis on anatomical features of the mouth, face, and neck that are known to affect both face-mask ventilation and endotracheal intubation (i.e., receding chin, small mouth opening, and short neck). Specifically, we assessed body weight as expressed by the body mass index (BMI), and anatomical variables, including mandibular retrognathia or prognathia, oropharyngeal volume (Mallampati modified score) (25), thyromental distance (TMD), mentohyoid distance (MHD), sternomental distance (SMD), mouth opening and interincisor gap (IIG), and, finally, history of medical or surgical disorders that could affect face-mask ventilation or endotracheal intubation, such as insulin-dependent diabetes mellitus (IDDM), cervical-neck stiffness, or thyroid surgery.
Each anatomical risk factor was stratified into a risk category, whereas clinical variables were assessed as present or absent. Preoperative risk factors included a BMI (BMI = weight [kg]/height [m]2) of >32 kg/m2, the presence of retrognathia or prognathia, a Mallampati scale score >2, TMD <6.5 cm, MHD <4.5 cm, SMD <12 cm, IIG <3.5 cm, a history of DI, the presence of IDDM for >5 yr, or the presence of cervical stiffness.
The oropharyngeal volume, as defined by the modified Mallampati scale (four scores), was assessed with the head in sniffing position, with the tongue protruded. Prognathia or retrognathia was assessed with a finger pressed against the mouth and chin directly in the midline under the nose (as indicating to be silent). Prognathia was identified as present if the finger angle was <90°, 90° was considered normal, and >90° was considered as retrognathia.
Mouth opening, defined as the IIG, was measured with the mouth fully opened without the use of any accessory muscles. Moreover, patients with a history of mandibular subluxation were asked to reach their known limit. In edentulous patients, the intergingival gap was measured. TMD, MHD, and SMD were measured along a straight line from the lower border of the mandibular mentum to the thyroid notch, the superior border of hyoid bone, and the sternal notch, respectively, with the head in sniffing position, avoiding full extension of the head by engaging accessory muscles.
The range of motility of the head on the neck was assessed with one finger on the mentum and one on the occipital spine and was evaluated as normal or reduced (cervical stiffness); if the finger on the occipital spine was higher or at the same level of the mentum with the head fully extended, motility was considered reduced.
The ability to perform bag mask ventilation (BMV) was evaluated after the induction of general anesthesia and paralysis. The ease of gas exchange during BMV was defined as difficult when one or more of the following factors were present: inability to maintain an adequate mask seal; inability to obtain chest excursion, obtain a good capnograph tracing, or maintain oxygen saturation >90% despite good muscle relaxation; the necessity of using a Guedel oral airway; or two-person BMV.
Anatomical risk predictors were also related to the difficulty of MV. BMV grades (easy/difficult) were compared with C-L simplified scale grades (1–2, easy; 3–4, difficult) and with the grades of an operator simplified difficulty scale (easy/difficult) for ease of ventilation. Laryngeal view was evaluated by direct laryngoscopy performed via a No. 4 Macintosh laryngoscope with the patient in the sniffing position according to the C-L scale scores (1–4).
The operator’s subjective judgment as to the ease of intubation performance was defined with the OIDS, as assessed by board-certified anesthesiologists with at least 5 yr of experience. Endotracheal intubation was rated as easy when immediate visualization of the glottis was obtained and tube progression into the trachea was achieved without the use of any intubation aid [rigid stylet, gum elastic bougie (26), or hollow intubator introducer (27)] within three attempts. Endotracheal intubation was rated as difficult in the presence of poor visualization of the glottis, when an intubation aid was used, or when three or more intubation attempts were required.
Linear correlation indexes were calculated for the Mallampati versus the C-L scale scores and for the simplified C-L (easy/difficult) versus the simplified OIDS (easy/difficult) grades. The significance of results regarding the grade of difficulty of endotracheal intubation or BMV, as well as any correlation between the simplified C-L and the simplified operator intubation and ventilation difficulty scale grades, were calculated by using χ2 analysis.
Furthermore, we determined the effect of the aforementioned preoperative risk factors in predicting the grade of difficulty of laryngeal visualization. True and false positives, true and false negatives (FN), sensitivity, specificity, PPV and negative predictive values (NPV), relative risks, and odds ratios (ORs) of each method of airway assessment were calculated.
Mallampati scale scores, anatomical features, medical risk factors and their relation to the C-L scale scores, as well as the OIDS scores, are shown in Table 1. No patients presented with a history of difficult airway (endotracheal intubation or BMV) or mandibular prognathia. Interestingly, no patient with a history of IDDM had a poor laryngeal view (C-L Grade 3 or 4).
An endotracheal intubation aid (rigid stylet, gum elastic bougie, or hollow intubating introducer) was used in combination with direct laryngoscopy to facilitate intubation in 72 patients. All patients were successfully intubated, and no patient was impossible to ventilate. Fiberoptic intubation was successfully accomplished in all patients with a C-L scale score of 4.
Table 2 reports the sensitivity and specificity of each risk factor in predicting difficult endotracheal intubation, defined as having a C-L scale score >3 or a “difficult” OIDS score. DI, as measured by the OIDS, occurred in 28 (1.4%) cases (6 patients, C-L Grade 2; 22 patients, C-L Grade 3), and 185 (9.4%) patients were considered difficult to intubate (13 patients, C-L Grade 1; 151 patients, C-L Grade 2; 21 patients, C-L Grade 3). The Mallampati versus C-L linear correlation (Fig. 2; scatterplot) index was 0.90 (P < 0.001). A Mallampati Class 3 correlated with a C-L Grade 2 (r = 0.94), whereas a Mallampati Class 4 correlated with a C-L Grade 3 (r = 0.85) and C-L Grade 4 (r = 0.80). The linear correlation of C-L versus OIDS was 0.96 (P < 0.001).
The use of an intubation aid compared with C-L groups of risk showed a sensitivity of 41.6%, a specificity of 99%, a PPV of 70%, and an NPV of 98% (P < 0.001). An intubation aid was used to accomplish endotracheal intubation in 55 (26%) patients categorized as difficult to intubate according to the OIDS scale score, compared with 27 (63%) patients graded as C-L Grade 3. Intubation aids were used in 17 cases of easy endotracheal intubation according to the OIDS, in comparison to 42 patients graded C-L Grade 1 or 2.
Only 47 patients (2.4%) exhibited DMV. Table 3 demonstrates the sensitivity and specificity of each anatomical feature in predicting difficult ventilation.
In our study, the incidence of DMV and DI and of impossible endotracheal intubation with direct laryngoscopy was less than previously reported in the literature (2,4–6,8). Our data demonstrate that the Mallampati score correlated with the C-L scale of laryngoscopic view, as suggested by Mallampati et al. (13) and confirmed by Samsoon and Young (25), Wilson et al. (16), and ourselves in a previous report (7), but it actually failed in this series to reach high sensitivity and PPV as a predictive risk factor for DI. Because DI is infrequent, there is a small incidence of FN, yet the sensitivity of tests should be high to detect possible difficulties with intubation.
Additionally, we found that the other anatomical landmarks and clinical risk factors were even less capable of reaching high sensitivity. The use of anatomical indexes associated with the Mallampati score did not result in any improvement of sensitivity and PPV. Indeed, Tse et al. (20) concluded that the Mallampati score, TMD, and cervical motility were of little value in predicting a difficult airway. Although the Mallampati scale and anatomical clinical risk factors did not show predictive properties in our series, they statistically increased the OR for prediction of a difficult airway.
Few studies have achieved significant predictive ability of individual or combinations of anatomical variables for difficult airway management. Multivariate airway risk indexes derived from large population samples appear to reduce the occurrence of FN and improve the PPV (6,8,18). According to our data, if single risk factors lack high sensitivity, they will affect the accuracy of screening.
Although an interesting intubation difficulty score has been developed by Adnet et al. (28) as an objective assessment of airway access difficulty, we prefer an operator-dependent subjective difficulty scale for training convenience. In our study, the OIDS and C-L scales of posttest predictive evaluation used to correlate clinical and anatomical risk factors of a difficult airway demonstrated good correlation (0.90; P < 0.001). Although this correlation was strong, we performed a finer evaluation of difficult endotracheal intubation by using airway adjuncts (e.g., intubation aids) as independent variables of a difficult airway, and we correlated the OIDS and C-L score with the use of those devices. We analyzed the use of these devices during laryngoscopy as a quality index of the real capability of the C-L score to characterize the cases of easy laryngoscopy but difficult to intubate and those of DL but easy to intubate, as defined by the OIDS.
The first issue considers difficult endotracheal intubation defined by the OIDS but previously considered easy by the C-L scale, usually presenting as C-L Grade 2. Although a poor laryngeal view is a determinant factor of DI, DI is not synonymous with restricted glottic visualization. Indeed, a large percentage of FNs for difficult endotracheal intubations classified as C-L Grade 2 were considered difficult cases according to the OIDS. A stylet or a bougie used in a C-L Grade 1–2 may imply an unpredicted average intubation and a reduced accuracy of the laryngoscopic view as compared with the OIDS.
The second aspect relates to difficult endotracheal intubation defined by a C-L Grade 3 but resulting in an easy OIDS where an aid was not used and intubation was simply solved by manipulations during the second attempt. Although this study was conducted under the supervision of skilled anesthesiologists, we could speculate a role of the conferred incidence of false positives at the C-L Grade 3 by unskilled individuals performing laryngoscopy.
Although the definition of DI is related to the concept of limited laryngoscopic view, we found that the laryngeal grade classification is not completely suitable to characterize an intubation. Revisions of the C-L classification of laryngeal inlet visualization (29,30) attempted to better estimate the chances that some C-L 2 grades could be average or difficult and that some C-L 3 grades could be easy.
The most dangerous situation occurs when intubation is impossible and mask ventilation is or becomes inadequate. DMV occurs more frequently in cases of DI; therefore, the prediction of DMV is of vital importance. Interestingly, we did not find any correlation between DMV and DI or DL and anatomical or clinical risk factors in this study. Further investigation of this matter is warranted.
In conclusion, ideal risk-assessment methods challenge high sensitivity and specificity, and predictive variables should balance the dynamic relationship between sensitivity and the NPV. Unfortunately, the Mallampati score and all of the other anatomical and clinical indexes had a low PPV in this study. Moreover, anatomical distances failed in predicting poor laryngeal view, possibly because they require standardization and correlation with the patient’s ethnicity, height, sex, and physical structure (tall, short, thin, plump). According to our data, the C-L scale requires revision, as suggested by Yentis and Lee (29) and Cook (30).
Achieving and maintaining a patent airway depends principally on anatomic and individual factors, but experience and alternative endotracheal intubation aids also play a significant role in these endeavors. For our daily practice, reducing the incidence of FN prediction is important, but preparedness for possible difficulty in airway management is absolutely necessary, and a difficult-airway cart with selected alternative airway adjuncts/devices should be readily available.
Even though preoperative evaluation is unable to provide objective clinical risk stratification, the clinical anesthesiologist makes daily assessments that are extremely important, and we believe that scores may represent a useful means for achieving a faster and more accurate prediction of a difficult airway for those unskilled in airway management.
We would like to thank Pietro Pietrini, MD, Maria Vagia Perri, Dawn Gralyn Iannucci, and Andrea Pescini, MD, for their assistance in the preparation of this manuscript.
1. Hovi-Viander M. Death associated with anaesthesia in Finland. Br J Anaesth 1980;52:483–9.
2. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia 1984;39:1105–11.
3. Rocke DA, Murray WB, Rout CC, et al. Relative risk analysis of factors associated with difficult intubation in obstetric anesthesia. Anesthesiology 1992;77:67–73.
4. Caplan RA, Posner KL, Ward RJ, Cheney FW. Adverse respiratory events in anesthesia: a closed claims analysis. Anesthesiology 1990;72:828–33.
5. Crosby ET, Cooper RM, Douglas MJ, et al. The unanticipated difficult airway with recommendations for management. Can J Anaesth 1998;45:757–76.
6. Arné 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;80:140–6.
7. Cattano D, Pescini A, Paolicchi A, Giunta F. Difficult intubation: an overview on a cohort of 1327 consecutive patients [abstract]. Minerva Anestesiol 2001;67(Suppl 2):45.
8. Langeron O, Masso E, Huraux C, et al. Prediction of difficult mask ventilation. Anesthesiology 2000;92:1229–36.
9. Rose DK, Cohen MM. The airway: problems and predictions in 18,500 patients. Can J Anaesth 1994;41:372–83.
10. Butler PJ, Dhara SS. Prediction of difficult laryngoscopy: an assessment of thyromental distance and Mallampati predictive tests. Anaesth Intensive Care 1992;20:139–42.
11. Frerk CM. Predicting difficult intubation. Anaesthesia 1991;46:1005–8.
12. Lewis M, Keramati S, Benumof JL, Berry CC. What is the best way to determine oropharyngeal classification and mandibular space length to predict difficult laryngoscopy? Anesthesiology 1994;81:69–75.
13. Mallampati SR, Gatt SP, Gugino LD, et al. A clinical sign to predict difficult tracheal intubation a prospective study. Can Anaesth Soc J 1985;32:429–34.
14. Oates JDL, Macleod AD, Oates PD, et al. Comparison of two methods for predicting difficult intubation. Br J Anaesth 1991;66:305–9.
15. Savva D. Prediction of difficult tracheal intubation. Br J Anaesth 1994;73:149–53.
16. Wilson ME, Spiegelhalter D, Robertson JA, Lesser P. Predicting difficult intubation. Br J Anaesth 1988;61:211–6.
17. Asai T, Koga K, Vaughan RS. Respiratory complications associated with tracheal intubation and extubation. Br J Anaesth 1998;80:767–75.
18. El-Ganzouri AR, McCarthy RJ, Tuman KJ, et al. Preoperative airway assessment: predictive value of a multivariate risk index. Anesth Analg 1997;84:419–21.
19. Turkan S, Ates Y, Cuhruk H, et al. Should we reevaluate the variables for predicting the difficult airway in anesthesiology? Anesth Analg 2002;94:1340–4.
20. Tse JC, Rimm EB, Hussain A. Predicting difficult endotracheal intubation in surgical patients scheduled for general anaesthesia: a prospective blind study. Anesth Analg 1995;81:254–8.
21. AARC (American Association for Respiratory Care) clinical practice guideline: management of airway emergencies. Respir Care 1995;40:749–60.
22. The ASA task force on guidelines for management of the difficult airway: practice guidelines for management of the difficult airway. Anesthesiology 1993;78:597–602.
23. Frova G. The difficult intubation and the problem of monitoring the adult airway: Italian Society of Anesthesia, Resuscitation, and Intensive Therapy (SIAARTI). Minerva Anestesiol 1998;64:361–71.
24. Boisson-Bertrand D, Bourgain JL, Camboulives J, et al. Difficult intubation: French Society of Anesthesia and Intensive Care—a collective expertise. Ann Fr Anesth Reanim 1996;15:207–14.
25. Samsoon GL, Young JR. Difficult tracheal intubation: a retrospective study. Anaesthesia 1987;42:487–90.
26. Kidd JF, Dyson A, Latto IP. Successful difficult intubation. Anaesthesia 1988;43:437–8.
27. Tuzzo DM, Frova G. Application of the self-inflating bulb to a hollow intubating introducer. Minerva Anestesiol 2001;67:127–32.
28. Adnet F, Borron SW, Racine SX, et al. The intubation difficulty scale (IDS). Anesthesiology 1997;87:1290–7.
29. Yentis SM, Lee DJ. Evaluation of an improved scoring system for grading of direct laryngoscopy. Anaesthesia 1998;53:1041–4.
© 2004 International Anesthesia Research Society
30. Cook TM. A new practical classification of laryngeal view. Anaesthesia 2000;55:274–9.