Definition, Incidence, and Predictors of the Difficult Airway

Arun L. Jayaraman

Paul E. Biegeleisen

The term “difficult airway” defies simple characterization as there is no published standard definition. It may be interpreted to indicate challenging or impossible mask ventilation (IMV), glottic visualization, and/or endotracheal tube placement. Regardless, prudence dictates careful study of these aspects of airway management. This is particularly true when considering that examination of the American Society of Anesthesiologists closed claims project indicates that a significant proportion of adverse anesthetic outcomes were associated with respiratory events, including inadequate ventilation, difficult tracheal intubation, and esophageal intubation, accounting for significant morbidity and mortality.¹ Of note, many of these events were deemed to be preventable. Similar observations were made in a retrospective analysis of over 80,000 anesthetics.² As such, the subsequent text contains a discussion of the definition, incidence, and predictors of difficult mask ventilation (DMV) and difficult intubation (DI) in adults.

DIFFICULT MASK VENTILATION

DMV has been defined using various criteria, including poor oxygenation as detected via pulse-oximetry, inadequate chest excursion, a leak around the mask, and necessity of two-handed mask ventilation. Previously, the American Society of Anesthesiologists Practice Guidelines for Management of the Difficult Airway defined DMV as the inability of an unassisted anesthesiologist to either maintain an SpO₂ > 90% using 100% oxygen and positive pressure mask ventilation (in a patient whose SpO₂ was greater than 90% prior to induction of anesthesia) or to prevent or reverse signs of inadequate ventilation during positive pressure mask ventilation.³ The updated guidelines provide a broader definition of DMV as the inability of an anesthesiologist to provide adequate mask ventilation due to poor mask seal, excessive gas leak, and/or significant impedance to gas entry or exit.⁴,⁵ Furthermore, these guidelines enumerate sundry signs of ineffective mask ventilation, including inadequate chest excursion, inadequate breath sounds, auscultatory evidence of obstruction, cyanosis, gastric insufflation, inadequate SpO₂, inadequate end-tidal carbon dioxide, inadequate spirometric measures of exhaled gas flow, and hemodynamic changes associated with hypoxemia and/or hypercapnia.

Despite the importance of adequate mask ventilation in managing the difficult airway, as emphasized in the American Society of Anesthesiologist’s difficult airway algorithm, research examining the incidence and predictors of DMV is somewhat sparse, particularly when compared with that regarding difficult tracheal intubation. An observational study involving 1,502 adults undergoing abdominal, gynecologic, orthopedic, urologic, and neurosurgery with general anesthesia indicates that DMV may be as common as 5%.⁶ The incidence of IMV was 0.07%. This study identified several independent risk factors for DMV/IMV: age > 55 years, BMI > 26 kg/m², presence of a beard, lack of teeth, and history of snoring. Interestingly, reduced mouth opening and a receding mandible, two indicators of DI which will be discussed in more detail subsequently, were not associated with DMV in a statistically significant fashion. Of concern, preoperative airway assessment only predicted DMV in 17% of patients with DMV.

In another observational study involving 22,660 adults undergoing general anesthesia, the authors observed an incidence of 1.4% for DMV and 0.16% for IMV,⁷ as assessed using a previously proposed numerical grading scale for mask ventilation.⁸ This study identified several independent risk factors for DMV and IMV. Specifically, predictors of DMV included age ≥ 57 years, BMI ≥ 30 kg/m², Mallampati III or IV classification, presence of a beard, severely limited jaw protrusion, and snoring. Predictors of IMV were a history of snoring and thyromental distance (TMD) less than 6 cm. Of note, a subsequent study by the same group involving 53,041 adults undergoing general anesthesia
identified several additional independent predictors of IMV, including male sex, history of obstructive sleep apnea (OSA), history of neck irradiation, Mallampati III or IV classification, and presence of a beard.⁹ Independent risk factors for difficult or impossible mask ventilation and DI were BMI≥30 kg/m², limited or severely limited mandibular protrusion, abnormal neck anatomy, history of snoring, and history of OSA.

Despite differences in the incidences of DMV and IMV described in published studies, ranging from 1.4% to 7.8% and 0.07% to 0.16%, respectively, which are likely attributable to varying definitions, common independent risk factors for DMV/IMV include increased age, increased BMI, presence of a beard, and snoring.⁶,⁷,⁹,¹⁰,¹¹ Although there is no single ideal means of prospectively screening patients for the likelihood of DMV/IMV, common sense indicates that one should consider a combination of factors that would maximize sensitivity, potentially at the expense of specificity. A reasonable set of criteria for predicting DMV/IMV may include history of DMV/IMV, abnormal neck anatomy/craniofacial abnormality, history of neck irradiation, male sex, increased age, increased BMI, history of snoring or OSA, Mallampati III or IV classification, limited jaw protrusion, lack of teeth, presence of a beard, poor atlanto-occipital extension, and pharyngeal pathology. Unfortunately, there is no robust prospectively validated screening algorithm for DMV/IMV. Having said that, one would anticipate the pretest probability of DMV/IMV increasing in proportion to the number of positive risk factors. The only readily modifiable risk factor is presence of a beard. Of interest, most published data indicate a strong correlation between difficult ventilation and intubation.⁶,⁹,¹⁰,¹² This is not surprising as many of the predictors for DMV/IMV also apply to DI, as detailed in the following text.

DIFFICULT INTUBATION

As with DMV, DI has been defined in various ways. In practice, DI frequently results from inability to obtain adequate glottic visualization with laryngoscopy. The American Society of Anesthesiologists Practice Guidelines for Management of the Difficult Airway defines difficult laryngoscopy as impossible visualization of any portion of the vocal cords following multiple attempts at conventional laryngoscopy. DI is characterized as requiring multiple attempts, whereas failed intubation is described as inability to properly place an endotracheal tube despite multiple attempts.⁴

A large prospective observational study involving 18,500 patients indicates incidences of difficult and failed intubation of 1.8% and 0.3%, respectively.¹³ This study showed a positive correlation between DI and obesity, decreased TMD, limited mouth opening, reduced neck extension, male sex, and poor laryngeal exposure.¹³ Overall, the incidence of difficult laryngoscopy, defined as Cormack-Lehane laryngoscopic view ≥ 3 (Fig. 9-1A-E), is 1% to 4%,whereas that of failed intubation is 0.05% to 0.35%.¹⁴ Of note, the probability of encountering IMV and impossible endotracheal intubation in the same patient is estimated at 0.0001% to 0.02%.¹⁴ The importance of predicting DI is underscored by a large retrospective study which found that almost half of anesthetic complications related to airway management were preventable, as they were thought to be a consequence of either failed recognition of DI or inappropriate choice of intubating technique.²

History of Difficult Intubation

History of DI is likely the most reliable predictor of future DI.¹⁵,¹⁶ Among the various prognostic factors for DI, history of DI is of particular value as it aids the clinician in managing the airway of a patient in whom intubation may not be predicted to be problematic by other measures. As such, the practice at our institution is to document DI in such a fashion that it is prominently displayed in the electronic medical record along with pertinent details and future recommendations; patients are also given a letter to show those that cannot access the electronic record. Of course, although a history of DI does not necessarily indicate future difficulty, prudence dictates a cautious and determinate approach to airway management in these patients. The corollary to this is that airway management may not remain facile in a given patient although it was previously documented as such.

Mallampati Classification

Including modifications, Mallampati scoring is the most widely used and studied preoperative airway examination tool, so much so that it is a standard component of the preoperative evaluation. Mallampati classification provides a qualitative estimate of tongue size relative to the oropharyngeal cavity, as the tongue must be displaced into the floor of the mouth in order to visualize the larynx during direct laryngoscopy. The Mallampati score is determined by the ability to visualize the uvula, faucial pillars, and/or soft palate.¹⁷,¹⁸ The original classification scheme comprised three categories of oropharyngeal classification that were found to correlate with glottic exposure during direct laryngoscopy in a statistically significant fashion: Mallampati I indicates visualization of the uvula, faucial pillars, and soft palate; in the Mallampati II classification the uvula is masked by the base of the tongue but the faucial pillars and soft palate remain visible; and Mallampati III denotes visualization of the soft palate only. Samsoon and Young¹⁹ subsequently modified the Mallampati scoring system, adding Mallampati IV, which refers to visualization of the hard palate only, as their retrospective analysis of 13 failed intubations linked the Mallampati IV classification to failed intubation. This modified Mallampati score (MMS) is depicted in Fig. 9-2A-E.

FIGURE 9-1 Cormack-Lehane grading of glottic exposure with direct laryngoscopy. A: Grades of laryngeal exposure.

(From Samsoon GL, Young JR. Difficult tracheal intubation: a retrospective study. Anesthesia. 1987;42:487-490 with permission.) B-E: Photographs of the various Cormack-Lehane grades, appear in order of increasing score — B: Grade 1 laryngeal exposure; C: Grade 2 laryngeal exposure; D: Grade 3 laryngeal exposure; E: Grade 4 laryngeal exposure.

Of note, as originally described, the Mallampati classification was assessed with the patient sitting upright and tongue maximally extended; head positioning and phonation were not specified.¹⁸ In developing the MMS, Samsoon and Young¹⁹ used the sitting position with the head neutral and tongue extended; phonation was not specified. A subsequent prospective analysis investigating the effects of patient positioning on oropharyngeal classification, which was in turn correlated with ease of direct laryngoscopy using the Cormack-Lehane system, indicates that ideal assessment occurs with the patient sitting with head extended, tongue maximally protruded, and phonation.²⁰ Subsequent comparisons of this extended Mallampati score (EMS) with MMS suggest that EMS is associated with 7% to 10% increased specificity, up to 83%, for difficult laryngoscopy and comparable sensitivity.²¹,²² Interestingly, there is some data indicating that MMS grade is increased by changing from the sitting to supine position.²³ Furthermore, MMS assessment done in the supine position may be associated with
increased positive predictive value for difficult laryngoscopy compared with that done in the sitting position.²⁴,²⁵ Fig. 9-2 contains photographs of the four classes in the modified Mallampati scheme and four grades of glottis exposure of the Cormack-Lehane system.