Chapter 10. Evaluation of the Patient with a Difficult Airway

1. A patient with a history of difficult intubation should be treated as having a difficult airway, even though physical appearance and physical examination may be unremarkable.

2. A patient with anatomic variations indicative of possible difficult intubation should receive a careful history and physical examination to define the extent of the potential airway problem.

3. Possible or potentially difficult intubation may be predicted by the Mallampati test, evidence of receding mandible, limited mouth opening as a result of tissue or temporomandibular joint (TMJ) restriction, enlarged teeth, high arched palate, narrow small mouth, or restricted cervical spine movement.

4. All current tests to predict difficulty with airway management are associated with a high incidence of false-positive and false-negative results and have low predictive value. To minimize airway-related complications, it is optimal to accept a high incidence of false-positive predictions by the various tests and treat any patient identified as having a possible difficult intubation accordingly.

5. Unexpected failed ventilation and intubation may result from oropharyngeal, laryngeal, or tracheal pathology that may not be identified by external examination.

6. In pediatrics, infection-related airway compromise and congenital airway malformations are the major airway management problems.

7. In adults, stridor at rest indicates a serious degree of obstruction with a cross-sectional opening less than 4 mm.

8. Upper airway endoscopy with a standard or videolaryngoscope and/or fiberoptic bronchoscope is useful in defining anatomic challenges in patients with upper airway pathology before induction of general anesthesia.

A challenging airway may present as difficulty with ventilation, difficulty with rigid laryngoscopic tracheal intubation, or both. The American Society of Anesthesiologists' (ASA) Practice Guidelines for Management of the Difficult Airway has defined difficult ventilation as a circumstance where "it is not possible for the unassisted anesthesiologist to prevent or reverse signs of inadequate ventilation during positive pressure ventilation."1

Difficult rigid laryngoscopy is defined as a situation in which "it is not possible to visualize any portion of the vocal cords with conventional laryngoscopy." A difficult intubation is defined as a circumstance in which "the proper insertion of an endotracheal tube using conventional laryngoscopy requires more than three attempts, or greater than 10 minutes."1

The incidence of difficult intubation by rigid laryngoscopy varies from 0.5% to 13.6% in published studies.2-8 Discrepancies in the reported incidence of difficult intubation are to be expected because most reports are retrospective studies and apply different definitions of what constitutes a difficult intubation.9 The incidence of failed intubation in the general surgical population has been reported as 1 in every 2230 patients.4 The incidence of failed intubation in the parturient population has been reported as 1 in 283 to 750 patients.4,8 This represents a 3- to 10-fold increase compared with the incidence in nonparturient patients. The precise frequency of difficult mask ventilation is unknown, but an Australian study indicated that 15% of difficult intubations were also associated with difficult mask ventilation.10 In an evaluation of 22,600 attempts at mask ventilation at the University of Michigan, 1.4% (313 cases) were difficult to ventilate, and 0.16% (37 cases) were impossible to ventilate. Of the difficult and impossible to mask ventilate cases, 0.37% (84 cases) were also observed to be a difficult intubation. Of the 37 impossible to ventilate cases, only 1 patient required surgical airway access.11

Approximately 30% to 50% of all anesthetic deaths have been attributed to the inability to manage a difficult airway.12 The ability to predict difficulty with ventilation and intubation accurately would help minimize disasters related to airway management. Unfortunately, the positive identification of all difficult airways is not currently possible. The coexistence of difficult ventilation is even more difficult to predict accurately.

Difficult airway management results from anatomic extremes or diseases affecting the airway. These factors may prevent a good mask fit, adequate positioning of a supraglottic airway, proper positioning of the head and neck, and opening of the mouth. Inability to ventilate the lungs or intubate the trachea may of course also result from poor technique and/or lack of technical expertise.13

Cormack and Lehane developed a classification for the view obtained at laryngoscopy.14 This classification uses four grades as follows: grade I, a full view of glottis; grade II, only the posterior commissure is visible; grade III, only the tip of the epiglottis is visible; grade IV, no glottic structures are visible (Fig. 10-1). Use of the original Cormack and Lehane scoring system led Yentis and Lee to develop a modified Cormack and Lehane scoring system in which grade II (only part of the glottis visible) was divided into IIa (part of the cords are visible) and IIb (only the arytenoids or the very posterior origin of the cords are visible) (Fig. 10-2).15 In this system, grade IIb denotes a laryngoscopic view that is relatively common and often associated with difficulty passing a tracheal tube. This system is now frequently used for recording the ease or difficulty in laryngoscopic view in the anesthetic record and in studies of tracheal intubation.

The Intubation Difficulty Scale (IDS) is another tool used in airway research that can be a useful indicator of total intubation difficulty in a prior intubation.16,17 The IDS is a quantitative measure of the total intubation difficulty encountered during a chosen procedure or sequence of procedures and is calculated after the fact. The score is based on 7 parameters known to be associated with difficult intubation. The 7 parameters are number of supplementary attempts, number of supplementary operators, number and type of alternative techniques used, laryngoscopic grade, subjective lifting force, the use of external laryngeal manipulation, and mobility or position of the vocal cords. The scoring of each individual parameter represents a divergence from ideal, and the total score represents the sum divergence from a zero difficulty, ideal intubation.

Patients with difficult tracheal intubation can be categorized into 3 groups (Box 10-1). The first group consists of patients in whom tracheal intubation is known or expected to be difficult. The second group includes patients with external anatomic features indicative of probable airway difficulties. The third group includes patients who present no external anatomic evidence or history of a difficult airway but nonetheless prove to be difficult.

These patients have anatomic variations that may interfere with rigid laryngoscopy and tracheal intubation. The failure to identify these patients correctly may lead to an improper anesthetic induction plan and an unexpected failed intubation. If ventilation proves impossible, a life-threatening emergency is created.

To identify patients with potentially difficult-to-manage airways, a careful history regarding patient breathing, sleep position, and voice quality is taken. During physical examination, the patient should be viewed from a frontal and profile view to assess mandibular size. Thyromental distance is measured, and neck rotation, flexion, and extension mobility are evaluated and graded. The neck is palpated for identification of the cricothyroid membrane. Additionally, one should check for TMJ problems, mouth opening, loose or protruding teeth, degree of overbite, size of the tongue, visibility of faucial structures, and patency of the nares.

A number of predictive tests have been described that detect potentially difficult intubation (Box 10-2). These clinical examinations are straightforward and can be done at the bedside. Although useful in airway evaluation, all these tests are associated with relatively high rates of false-positive and false-negative predictions (Box 10-3).

Anatomic Considerations

In the early development of rigid laryngoscopy, certain anatomic features were recognized as contributing to the difficulty of intubation.18 These features included a recessed mandible, limited extension of the head and neck, limited mouth opening, and mandibular depth.

Five physical features were evaluated prospectively by Wilson et al in an attempt to identify potentially difficult intubations (Table 10-1).6 Each risk factor was given three possible scores (0, 1, or 2). A total score of greater than 2 predicted a 75% chance of difficult intubation but with a significant incidence of false-positive results. Changing the prediction criteria to a score of greater than 4 reduced the number of false-positive results but also increased the incidence of false-negative predictions. In this study of 778 patients, 1.5% of rigid laryngoscopies were found to be difficult.

Atlantooccipital Mobility

Adequate cervical mobility and mouth opening are essential for alignment of the oral, pharyngeal, and laryngeal axes required for visualization of the glottis during rigid laryngoscopy.18,19 Decreased neck mobility also limits maneuvers that can be used to keep the airway open, thereby predisposing to difficult mask ventilation.

Evaluation of atlantooccipital extension is performed by having the patient sit straight and extend the head while maintaining the cervical spine in a neutral position.19 The greater the atlantooccipital distance in the neutral position, the greater the possible degree of head extension. A reduction in atlantooccipital extension by a third or more will contribute to the difficulty of intubation.19 If the posterior tubercle of the atlas is in contact with the occiput in the neutral position, attempts to extend the head result in anterior bowing of the cervical spine and forward displacement of the larynx.20 Limitations in head extension may occur secondary to anatomic variations of the atlantooccipital gap in otherwise healthy people or secondary to pathologic conditions such as rheumatoid arthritis.

Mouth Opening, Dentition, Mandibular Space

In addition to head and neck extension, other anatomic factors may interfere with the line of vision from the mouth opening to the vocal cords. The hinge movement of the mandible controls mouth opening. A horizontal gliding movement allows for subluxation of the mandible, which allows additional anterior displacement of the tongue during rigid laryngoscopy. A mouth opening (distance between mandibular and maxillary central incisors) limited to 3.5 cm or less tends to make intubation more difficult. TMJ dysfunction, congenital fusion of the joints, trauma, tissue contracture around the mouth, and trismus may limit mouth opening. Trismus secondary to pain associated with infection usually relaxes under general anesthesia, but this cannot be assured if the infection has been long standing or involves the pterygoid space.

In addition, protruding maxillary anterior teeth may interfere with laryngoscope placement and passage of the endotracheal tube during intubation.

During rigid laryngoscopy, the tongue is displaced into the mandibular space, opening the line of vision from mouth to the larynx. A small mandibular space may fail to accommodate tongue displacement adequately, thus interfering with visualization of the larynx.

Thyromental Distance

Patil et al reported that rigid laryngoscopy may be impossible in adults if the thyromental distance (TMD) is less than 6.0 cm (3 finger breadths).21 Practitioners are cautioned that when measuring TMD, the use of a paper ruler is preferred over the traditional thought that 3 finger breadths equal 6.0 cm unless one has checked which distal interphalangeal joint combination equals 6 cm. This distance may be 3 fingers on those with larger hands and 4 fingers on those with narrow fingers.22 Alignment is more difficult in the patient with a receding mandible where the TMD short, as the laryngeal axis forms a more acute angle with the pharyngeal axis. The TMD is measured between the bony point of the mentum of the mandible and the thyroid notch with the head fully extended. Maximum extension of the head ensures reproducibility of measurements.

Sternomental Distance

The sternomental distance is measured with the head fully extended and the mouth closed. This measurement is reported to be more sensitive and specific in predicting difficult rigid laryngoscopy than the Mallampati test, TMD, mouth opening, and mandibular subluxation measurements.5 The predictive value of sternomental distance regarding difficult intubation has not been studied by other investigators.

Visibility of Oropharyngeal Structures

Predicting probability of difficult laryngoscopy by physical evaluation was reported by Skolimowski et al in 1975 when describing a micrognathic patient who could not be intubated: "The anteroposterior dimension of the pharynx was markedly reduced by a large tongue reaching far posteriorly. It was not possible to examine the lower pharynx by means of a laryngoscope mirror because the tongue was situated too far posteriorly to be pulled forward."23

Mallampati described the examination signs and related them to intubation difficulty. He correlated the degree of visibility of the oropharyngeal structures with the difficulty of rigid laryngoscopy.24 In this study, a sitting patient was asked to open his or her mouth as wide as possible and maximally protrude the tongue. The visibility of the faucial pillars, soft palate, and uvula was noted. The airway was classified into 3 categories: class I, soft palate, fauces, uvula, and pillars are visualized; class II, soft palate, fauces, and pillars are visualized, but the uvula is masked by the base of the tongue; and class III, only the soft palate can be visualized. In class III, visualization of the glottis with rigid laryngoscopy is expected to be difficult.24 Samsoon and Young extended the oropharyngeal exposures to include a fourth class.4 This four-category system is in common use and classified as follows: class I, soft palate, fauces, uvula, and pillars are visualized; class II, soft palate, fauces, uvula are seen; class III, only the soft palate and base of the uvula are observed; class IV, the soft palate is not visible (Fig. 10-3). A further modification of the Mallampati visualization scoring included a class 0 view.25,26 Class 0 is defined as the ability to see any part of the epiglottis upon mouth opening and tongue protrusion. Ezri et al evaluated 764 patients and reported that 1.18% of patients had a class zero airway.27 A class 0 airway was noted to be an excellent predictor of uncomplicated laryngoscopy.

In a retrospective study of 13 patients with failed intubations, Samsoon and Young found that in 12 of the patients there was a good correlation between the degree of difficulty of tracheal intubation and the visibility of the oropharyngeal structures.4 Rocke et al reported the association between Mallampati class and difficulty of intubation to be significant (p <0.001), but only 6.6% of class IV airway cases were associated with difficult tracheal intubation, and all could be intubated.8 Using the concept of relative risk, they calculated the probability of difficult intubation. The presence of class III airway was associated with a relative risk of 7.58 times greater than class I, and airway class IV had a relative risk of 11.2 times greater than class I. Oates et al compared Mallampati's classification with that of Wilson's 5 risk factors.28 The predictive value of both tests was found to be similar, but both tests had a high incidence of false-positive and false-negative predictors. Interobserver variability was less with Wilson's scoring system. Other investigators have found Mallampati's classification too subjective and the class III classification not useful as a predictor of difficult intubation when used alone.29-31

Phonation during tongue protrusion increases the specificity of the Mallampati test, thus improving its correlation with difficult intubation, but also increases the number of false-negative results.30 Tham et al reported that phonation created a marked improvement in view.32 Evaluation of the airway in the supine position produced a small but nonsignificant worsening of the view. Lewis et al recommended that the visibility of oropharyngeal tissues be conducted with the patient in the sitting position, with the head in full extension, the tongue out, and during phonation.33

In a retrospective review of failed intubations by Samsoon and Young, 7 of 1980 obstetric patients could not be intubated (incidence of 1 in 280) and 6 of 13,380 surgical patients could not be intubated (incidence of 1 in 2230).4 All patients had a class IV airway except one class II patient who had tracheal stenosis. These authors believed that preanesthetic assessment using the Mallampati classification had merit. Lee et al used the Cormack and Lehane grading system in their systematic meta-analysis review of the accuracy of Mallampati tests to predict the difficult airway.34 In their review of 42 studies involving 34513 patients, they found that both versions of the Mallampati test had good accuracy for predicting difficult laryngoscopy. The modified Mallampati score had good accuracy, but the original Mallampati had poor accuracy in predicting difficult tracheal intubation. Both versions of Mallampati score were poor at identifying difficult mask ventilation. The authors concluded that to be useful the Mallampati test must be part of a comprehensive evaluation that also includes assessment of dentition, TMD, and neck extension, as recommended by the ASA Task Force on the management of the difficult airway.1

In summary, the Mallampati test has been broadly applied, even though its predictive value for difficult intubation is low. As a result, some essential factors, such as suppleness and mobility of the neck or dentition status, may be minimized. No single test is perfect, but this should not discourage practitioners from seeking a more reliable test or combination of tests, especially as most currently used tests require only a minimal amount of time during the physical examination.35

Anteriorly Tilted Larynx

Roberts et al measured the degree of anterior tilt of the thyroid cartilage relative to the horizontal and demonstrated a relationship between degree of the thyroid cartilage tilt and the difficulty of laryngeal exposure using a Macintosh laryngoscope.36 Laryngeal tilt can be directly measured using a bubble inclinometer. When the tilt of the anterior surface of the thyroid cartilage is greater than 40°, rigid laryngoscopy was predicted to be difficult. The sensitivity of this test was reported to be 70%, with a specificity of 95% and a positive predictive value of 80%. Anterior laryngeal tilt is reduced by depression of the thyroid cartilage but is increased by cricoid pressure.

Direct Visualization Assessment

Upper airway endoscopy preoperatively using topical anesthesia for patient comfort provides excellent information in defining anatomic challenges in patients with upper airway pathology. The fiberoptic nasopharyngeal scope and the pediatric bronchoscope have been the standard for this evaluation. The classic rigid laryngoscope has not been as useful due to patient discomfort when it is used. However, newer model videolaryngoscopes may be used due to their ability to visualize the oropharyngeal and supraglottic regions more comfortably.

Radiographic Assessment

Lateral radiographs of the head and neck and the measurement of distances between bony landmarks have been used to identify predictive factors for difficult rigid laryngoscopy.37 Reduction in the distance between the occiput and the spinous process of C1 (atlantooccipital distance), and to the C1-C2 interspinous gap, has been correlated with the degree of difficulty in intubation. Reduction in the atlantooccipital gap may be found in some otherwise normal cases. Any limitation of cervical extension may interfere with rigid laryngoscopy and glottic exposure. Radiographic assessment is not considered cost-effective for routine airway evaluation.

Multiple Factors

Combining several tests for the prediction of difficult laryngoscopy and tracheal intubation improves the accuracy of the assessment. The simplified airway risk index (SARI) was developed to evaluate airway status preoperatively using multiple factors: mouth opening measurement, TMD, ability to protrude the mandible, Mallampati class, head mobility, and body weight.38 Calculation of the SARI in 136 patients revealed good interobserver agreement in assessing the Mallampati classification, mouth opening, and mandible protrusion. However, because there are more variables in measuring the TMD and evaluating neck mobility, there was not good interobserver agreement with these factors.39

Multiple factors often play roles in both difficulty with intubation and difficulty with mask ventilation. In a prospective study of 1502 patients, difficult mask ventilation was reported in 5%, with one occurrence of impossible ventilation.40 Difficulty with mask ventilation was anticipated by the anesthesiologist in only 17% of the cases in which difficulty occurred. Using multivariate analysis, 5 criteria were recognized as independent factors for difficult mask ventilation: age older than 55 years, body mass index (BMI) greater than 26 kg/m2, presence of a beard, lack of teeth, and a history of snoring. The presence of two factors suggested a high likelihood of difficult mask ventilation.

Bellhouse and Dore reported that combining factors (eg, limited atlantooccipital extension, chin protrusion, and tongue size) increases the sensitivity of predicting a difficult intubation.19 Frerk reported that combining two tests, Mallampati and TMD, improved the specificity of the prediction to 97.8% but did not improve sensitivity, which remained at 81.2%.41 A meta-analysis performed by Shiga et al of preanesthetic airway evaluation test performance on 50760 patients in 35 studies demonstrated a 5.8% overall incidence of difficult intubation.42 Each test in their analysis, including the Mallampati classification, TMD, sternomental distance, mouth opening, and the Wilson risk score,6 possessed only poor to moderate sensitivity and moderate to fair specificity. The most useful evaluations were found to be combinations of the Mallampati classification and TMD distance. These authors concluded that the clinical value of airway evaluation tests for predicting difficult intubation remains limited.

A stepwise approach to decision making in the evaluation of the airway is the airway approach algorithm.43 This algorithm is based on 5 clinical questions: Is airway control necessary? Is there potential for difficult laryngoscopy? Can supralaryngeal ventilation be used? Is there an aspiration risk? Will the patient tolerate an apneic period? Answers to these basic questions can help guide the practitioner in potential use of the ASA difficult airway algorithm.

Patients with a history of difficult ventilation or intubation and patients with anatomic or abnormal conditions associated with a complex airway fall into the category of known difficult airway.44 The causes of the expected difficult airway may be grouped into congenital or acquired conditions and be further classified on the basis of the location of involvement or disease.

Increased frequency of ventilation, chest retractions, increased use of accessory muscles, stridor, voice weakness, or hoarseness, alone or in combination, may indicate a potential airway problem (Table 10-2). Stridor is a particularly important sign and may provide evidence of the site and severity of airway obstruction related to severe oropharyngeal, glottic, and/or upper tracheal occlusion. Stridor during inspiration generally indicates obstruction at or above the larynx. Expiratory stridor is most often associated with intrathoracic or subglottic obstructions. Obstruction associated with the larynx or glottic region may produce biphasic stridor, although either inspiratory or expiratory sounds may predominate. In the adult, stridor at rest indicates a serious degree of obstruction with a cross-sectional airway opening of less than 4 mm or an irregularly narrowed airway several centimeters in length.

Obesity

Airway assessment of the obese patient should be performed with the patient in both the sitting and supine positions. Respiratory function and airway patency can be significantly altered by this change in position.45 A large neck circumference is associated with obstructive sleep apnea (OSA) in obese patients. In evaluating 123 patients with thick necks for OSA, Katz et al found that the sleep apnea-hypopnea index correlated with external neck circumference, BMI, and the internal circumference of the distal pharynx.46 Men more commonly have sleep-disordered breathing than women, and the sleep-disordered breathing tends to be more severe.47 In an evaluation of 3942 OSA patients, the frequency and severity of OSA in the sleep clinic population was found to be greater in men than women, with unknown factors other than neck circumference, age, and BMI contributing to the gender differences.48 In a prospective study of 100 morbidly obese patients (BMI >40 kg/m2), preoperative measurements of height, weight, neck circumference, width of mouth opening, sternomental distance, TMD, and Mallampati score were recorded.49 The view during direct laryngoscopy was graded, and the number of attempts at tracheal intubation was recorded. Neither absolute obesity nor BMI was associated with intubation difficulties. Large neck circumference and high Mallampati score were the only predictors of potential intubation problems.

In the supine position, changes in chest compliance and vital capacity may interfere with adequate spontaneous ventilation. The incidence of hiatal hernia, gastric pH of 2.5 or lower, and reduced functional residual capacity found in obese patients places these patients at increased risk for the consequences of aspiration of gastric contents.50,51 To minimize the risk of aspiration, a rapid sequence induction is commonly performed in obese patients.

There is consensus that airway management is more difficult in the morbidly obese patients. Opinions differ, however, on the difficulty of endotracheal intubation. Wilson and coworkers regarded obesity as a weak predictor of difficult intubation.6 Buckley et al reported a 13% rate of difficult intubation using a rapid sequence technique in the obese patient.51 Rocke et al excluded obesity as a risk factor in intubation.8 Bond found no correlation between BMI and difficulty of laryngoscopy.50 Juvin et al compared difficulty in tracheal intubation in obese to lean patients using the IDS and patient vital signs.52 A Mallampati score of III-IV was the only independent risk factor for difficult intubation in obese patients. Difficult tracheal intubation was more frequent in obese (15.5%) than lean (2.2%) patients. The use of the IDS score demonstrated that tracheal intubation, not laryngoscopy, was more difficult in obese than lean patients.

Body weight may not be as critical as the location of excess weight. Massive weight in the lower abdomen and hip area may be less important than when the weight is in the upper body area. A short, thick, immobile neck caused by cervical spine fat pads will interfere with rigid laryngoscopy. Furthermore, the redundancy of soft tissue structures inside the oropharyngeal and supralaryngeal area may also make visualization of the laryngeal structures difficult.

Mask ventilation may prove difficult in the obese patient. When high positive pressure is required to ventilate the patient, the chance of inflating the stomach is increased. Rapid oxygen desaturation during apnea, secondary to reduced functional residual capacity, limits available intubation time. In the case of the cannot intubate, cannot ventilate situation, access to the neck for transtracheal jet ventilation or establishing a surgical airway (eg, emergency tracheostomy or cricothyroidotomy) will also be more complex.

Pregnancy

The incidence of failed intubation is predicted as 1 in 300 patients undergoing cesarean delivery.4 Airway-related problems account for a third of all anesthetic-related maternal mortality.53 During pregnancy, mucosal vascular engorgement, laryngeal edema, immobility of the floor of the mouth related to tongue engorgement, enlarged breasts, and general weight gain contribute to difficult intubation.53-55

Airway anatomy may become distorted during prolonged labor or toxemia, leading to edematous soft tissue encroachment of the upper airway.54,55 Nasal intubation in these patients should be avoided because the mucous membranes become increasingly engorged and friable during late pregnancy. Similar to the obese patient, the obstetric patient should be considered to have a full stomach and at increased risk for gastric aspiration.

The physical changes created by pregnancy may lead to marked alteration in cardiovascular and respiratory function when the patient changes from a sitting to a supine position. Furthermore, in cases of fetal distress or maternal hemorrhage, the emergency nature of the circumstances compounds airway management problems.

Rheumatoid Arthritis

The airway management of these patients should be based on an understanding of the pathologic changes affecting the airway. In patients with advanced rheumatoid arthritis and spondylosis, airway management may be extremely difficult. Rheumatoid arthritis may involve any joint of the body, including the cervical spine, TMJ, and cricoarytenoid joint. A change in voice, the presence of dysphagia, dysarthria, stridor, or a sense of fullness in the oropharynx may indicate laryngeal involvement. A careful fiberoptic examination of the larynx and glottic structures may be informative when such signs and symptoms are present. An edematous larynx with hyperemic arytenoids and/or mucosa with swollen aryepiglottic folds and false cords may be observed. Changes in phonation may be associated with decreased mobility of the vocal cords. In the case of a narrowed glottic opening, endotracheal intubation frequently requires a smaller sized endotracheal tube. TMJ ankylosis may prevent orotracheal intubation because of limited mouth opening.

Physical examination of the patient with rheumatoid arthritis should include flexion, extension, and rotation of the head with palpation of the larynx and trachea for evidence of deviation and/or limitation. Upper-extremity radiculopathy suggests cervical spine arthritis. Progressive cervical spondylosis associated with rheumatoid arthritis leads to severe flexion deformity of the cervical spine, which complicates airway management.56 Synovial destruction and vertebral erosion, along with ligamentous changes, lead to instability of the cervical spine.57 Instability of the atlas and odontoid or of subaxial vertebral alignments may lead to subluxation of the cervical spine and cord compression. Cervical spine flexion and extension radiographs may be required for evaluation of instability and potential spinal cord compression. Although chin lift and jaw thrust are commonly used to improve mask ventilation and oxygenation, these maneuvers may increase the possibility of spinal cord compression and damage.58 If a head and neck stabilizing device is used by the patient, it generally should be left in place to prevent unintended movement of the cervical spine.59 Chest wall distortion in patients with rheumatoid arthritis may produce a major decrease in total lung volume and vital capacity. Pulmonary function tests may be helpful in determining a patient's ventilatory status in some cases.

Congenital Disease

Anomalies of the cardiovascular, nervous, musculocutaneous, endocrine, or excretory systems may produce abnormalities of the head, neck, or upper airway. Rosenberg and Rosenberg have tabulated the syndromes most often accompanied by aberrations of the upper airway.44 These include Crouzon, Goldenhar, Pierre Robin, and Treacher Collins syndromes, which are known for their grossly abnormal head and neck anatomy. Patients with congenital malformations associated with micrognathia, retrognathia, and macroglossia have a smaller oropharyngeal cross section and are prone to soft-tissue upper-airway obstruction.54,60 Children with craniocarpotarsal dysplasia have severe microstomia that becomes more inadequate as they grow older and develop teeth. These children often require repeated anesthetics for correction of their musculoskeletal and soft tissue deformities and can pose a significant problem for the anesthesiologist.

The most significant vascular malformations related to airway compromise are vascular rings, usually of aortic arch origin, encircling the trachea. Tracheomalacia, congenital tracheal stenosis, shortened trachea, and bronchogenic cysts can contribute to difficult airway management (Fig. 10-4).61 Wells et al reported that a significant percentage of infants with congenital malformation syndromes associated with cardiovascular anomalies and skeletal dysplasia have a shortened trachea.62 These infants may benefit from fiberoptic evaluation of endotracheal tube position to avoid unrecognized bronchial intubation.

Acromegaly, the syndrome that results from the pituitary gland producing excess growth hormone after epiphyseal plate closure at puberty, requires careful airway evaluation. This condition is seldom seen today in the United States due to early diagnosis and management. Most patients with acromegaly have poor Mallampati grade due to soft tissue overgrowth and macroglossia. The practitioner may also wish to include evaluation of growth hormone levels and duration of disease symptoms in evaluation of cases with this condition.63,64 Congenital malformation syndromes also may be associated with varying degrees of acute, progressive, or chronic airway obstruction. Congenital tumors or cysts may invade or obstruct the airway. Preoperative assessment should include determination of the site of the tumor and the extent of obstruction or distortion of the airway.

Airway Infection

Inflammation and edema can distort anatomy, fix soft tissues, and compress the airway, interfering with ventilation and intubation.65,66 Airway compromise by infection poses a major airway management problem in patients younger than 10 years. Of 90 deaths resulting from upper airway obstruction in children, 36 were related to airway infections (Fig. 10-5).67,68 Anesthetists are most often involved in the management of urgent conditions such as peritonsillar abscess, retropharyngeal abscess, submandibular abscess, Ludwig angina, croup, and epiglottitis. Each of these infections presents in a specific manner, which then dictates airway management.

Peritonsillar Space Infections

The peritonsillar space is a potential space located at the junction of the oral cavity and oropharynx. It is formed by the palatine tonsil, as well as the palatoglossus, palatopharyngeal, and superior pharyngeal constrictor muscles. Clinical findings of peritonsillar abscess include acute onset of fever, pain, dysphagia, and cervical adenopathy. As the infection spreads, it may involve the muscles of mastication, producing trismus secondary to pain and spasm. On examination, there is displacement of the uvula to the contralateral side, tonsillar enlargement, and fetid breath. At times, surgical drainage of the peritonsillar space is indicated. Management of the airway in this subset of patients is generally accomplished with orotracheal intubation. Rapid sequence intubation is generally possible. Reduced interdental distance that may be noted on preoperative evaluation of these patients is due to pain and usually resolves on administering anesthetic agents.

Retropharyngeal Space Infections/Prevertebral Space Infections

The retropharyngeal space is a midline compartment located between the middle and deep cervical fascia. The prevertebral space is a bilateral space located posterior to the deep cervical fascia. Infection of the retropharyngeal space most commonly occurs in children and presents with irritability, fever, dysphagia, muffled speech or cry, noisy breathing, stiff neck, and cervical adenopathy.69 Prevertebral space infection is much less common and is seen after spread of retropharyngeal abscess or more rarely primary infection of the prevertebral space. Airway management in these infections typically uses endotracheal intubation or tracheotomy. During laryngoscopy after induction, the posterior pharyngeal wall appears displaced anteriorly. Care must be taken to avoid lacerating and draining a posterior pharyngeal wall abscess with subsequent aspiration before the airway is controlled.

Ludwig Angina/Submandibular Space

Infection that arises in this space may be odontogenic in origin or due to chronic sialadenitis. The submandibular space is divided by the mylohyoid muscle. Infection localized above the mylohyoid muscle creates edema and distortion of the floor of the mouth. Upper airway endoscopy with a standard or videolaryngoscope and/or fiberoptic bronchoscope is useful in defining anatomic challenges in patients with upper airway pathology before induction of general anesthesia.

An infection inferior to the mylohyoid muscle may displace the base of tongue posteriorly. Ludwig angina is typically seen in patient with poor oral hygiene and presents as a bilateral neck cellulitis with displacement and swelling of the tongue base and floor of mouth. (Figs. 10-6 and 10-7). Both submandibular space infection and Ludwig angina present with acute onset of fever, dysphagia, pain, and swelling. The airway is best managed in these cases in a collaborative manner with the surgical team.

Epiglottitis

With the advent of the haemophilus influenza B vaccine in 1991, the incidence of epiglottitis has precipitously decreased by 90%. Although uncommon today, acute epiglottitis does occasionally present, particularly in the pediatric population, with a rapid onset of high fever, respiratory distress, drooling, and painful swallowing. On examination patients are noted to have tachycardia, tachypnea, and appear toxic. On phonation, a muffled quality may be appreciated.70 This classic presentation was often seen in the pediatric population before the introduction of the Hib vaccine. In the adult population, acute epiglottitis may be preceded by a viral upper respiratory infection. Noninfectious causes of epiglottitis may also be seen after thermal injury or caustic ingestion. Thermal injury after crack cocaine abuse is not infrequent. Once again, the airway is best managed in these cases in a collaborative manner with the surgical team. Any manipulation of the airway prior to intubation should generally be avoided.

Croup

Acute croup, or laryngotracheobronchitis, is often caused by the parainfluenza virus in children between the ages of 1 and 3, particularly in the spring and fall. Children typically present with low-grade fever, increased respiratory rate, barky cough, and hoarseness.70 Most children can be managed without intubation. When retraction or oxygen desaturation is noted, orotracheal intubation is indicated with an appropriate-size endotracheal tube.

Trauma

Trauma to the head and neck may produce major acute or chronic anatomic changes. These changes may affect airway accessibility, making tracheal intubation or mask ventilation difficult. Blunt or penetrating trauma to the larynx, trachea, hyoid structure, and facial bones can result in a complex, difficult-to-manage airway.71,72 Subcutaneous emphysema, hoarseness, stridor, and tracheal deviation are warning signs of airway injury. Such patients should be observed closely because progression of the condition may lead to airway obstruction.

The signs and symptoms of laryngeal trauma may be quite subtle. Patients with laryngeal trauma are often hoarse or short of breath, although this clinical presentation may not correlate with the severity of injury. Dysphagia is not a common symptom of laryngotracheal injury. Nevertheless, esophageal injury should be strongly considered in patients with laryngotracheal trauma. On physical examination, the presence of hemoptysis may indicate laryngeal or tracheal injury. External palpation of the neck should include evaluation of the hyoid bone, and thyroid and cricoid cartilages (Fig. 10-8). The skin should be examined for abrasions and subcutaneous air. Open wounds should not be probed, but entrance and exit wounds should be noted to better understand the trajectory of the injury, particularly with bullet wounds. Cricotracheal separation should be suspected when the mechanism of injury is via a "clothesline" or hanging injury. In these patients, stridor and subcutaneous emphysema should prompt immediate evaluation. Orotracheal intubation is contraindicated because it may cause more harm than good. In these cases, an endotracheal tube can migrate through a perforation into the cervical soft tissues or mediastinum, creating a tenuous and possibly dangerous airway for the patient.73 Thus awake tracheostomy below the site of injury remains the mainstay of airway management in these cases.

The trauma patient should also be examined for cervical spine injuries because movement of the neck during intubation may lead to irreversible paralysis. Maintaining cervical collar placement or applying axial traction may minimize spinal cord injury during intubation.74 The anesthetist should also be aware that mouth opening may be limited in the patient with facial trauma. Improvement in the ability to open the patient's mouth after induction of anesthesia and paralysis cannot be guaranteed. Therefore, a fiberoptic bronchoscope and a tracheostomy tray should be available for awake management of the airway.

Tumors

Head and neck tumors, both benign and malignant, may make intubation difficult. Mouth opening and proper positioning of the head and neck for rigid laryngoscopy can be limited by tumors, surgical scars, or radiation fibrosis of head and neck tissues (Table 10-3).2,57

Tumors of the oral cavity and oropharynx may create trismus due to invasion of the muscles of mastication. This trismus can often be overcome after induction with muscle relaxants. Anesthesiologists should work collaboratively with the otolaryngology service to perform safe laryngoscopy for intubation, diagnosis, and staging biopsies. Care should be taken to avoid trauma of these tumors, which will cause bleeding. This is especially true of friable tumors of the base of tongue and tonsil that, in addition to trismus, prevent proper mask ventilation and preclude the use of a supraglottic airway.

Tumors of the larynx create significant difficulty during orotracheal intubation. Exophytic tumors above the vocal cords may prolapse into and block the airway when even the slightest bit of sedation is administered.75,76 Once again, consultation with the otolaryngology service is indicated to avoid an emergently obstructed upper airway. In these cases, the airway can be managed with closed laryngoscopes such as a Dedo or anterior commissure (Holinger) laryngoscope (Fig. 10-9). Preparation for an awake tracheostomy under local anesthetic may be required in this subset of patients.

The presence of cancerous goiters is also a concern. Difficult tracheal intubation is reported in 17 (5.3%) of 320 patients undergoing thyroidectomy.77 In this study, multivariate analysis suggests that the presence of a cancerous goiter and Cormack and Lehane grade III or IV laryngoscopic view are independently associated with difficult intubation. However, the same study reported that the size of goiter was not associated with increased difficulty with intubation. In any case, it is important to be aware that a goiter can narrow the airway due to extrinsic compression (Fig.10-10). Last, although infrequent, vocal cord paralysis due to malignant thyroid disease or due to surgical trauma may complicate extubation.

Radiologic studies are indicated in the presence of trauma or tumors in or near the airway.78 Lateral cervical spine films, computed tomography (CT), or magnetic resonance imaging (MRI) may be used to assess the degree of airway compression and the involvement of associated structures. Topical anesthesia with fiberoptic laryngoscopy and bronchoscopy may prove beneficial in airway inspection.59

Airway Stents

The purpose of an airway stent is to prevent obstruction caused by malacia, stricture, or extrinsic compression that is not suitable for surgical correction either due to location or morbidity (Figs. 10-11 and 10-12).79,80 Conditions treated by stent placement include tracheomalacia, postintubation stricture, stricture related to lobectomy, tuberculosis, traumatic injury or compression secondary to malignancy, multinodular goiter, or an intrathoracic process. Tracheostomy and Montgomery T tubes are used extensively in the management of glottic, subglottic, or tracheal stenosis due to benign or malignant disease. More recently, tracheobronchial stents have been developed to help manage malacia, extrinsic compression, and stenosis of the distal airway (Figs. 10-13 and 10-14).

Preoperative evaluation requires communication with the otolaryngologist to verify the underlying diagnosis, position of the stent, and the best method of airway management. Presence of a stent may complicate or prevent traditional endotracheal intubation. Also, attempted orotracheal intubation with a stent in place can damage the airway or dislodge the stent. Laryngeal mask airways, however, can often be used to safely manage the airways of such patients.

Frequently, laryngectomy patients wear a silastic stent or a heat-moisture exchange system over the stoma. These "lary" tubes or buttons are uncuffed and often used in conjunction with a speaking valve or prosthesis placed across the tracheoesophageal wall (Figs. 10-15, 10-16, and 10-17). Given that these patients are "neck breathers" with potentially confusing surgically altered anatomy, their airway is best managed by collaborating with an otolaryngologist. Typically direct intubation of the stoma after removal of the laryngectomy tube is the best method of airway management. The speaking valve should not be removed during intubation because this is an indwelling device and will not obstruct placement of a tracheal tube.

Intrathoracic Lesions

Intrathoracic lesions can compromise airway integrity through compression of the tracheobronchial tree or by invasion of the trachea or bronchi. Mediastinal lesions leading to life-threatening airway obstructions may be found in neonates, infants, children, or adults.66,81-84 Congenital tumors or tumors arising in early infancy include hemangiomas, lymphangiomas, cystic hygromas, teratomas, dermoids, rhabdomyosarcomas, neurofibromas, neuromas, and thymic hyperplasia. Adults with mediastinal masses, commonly lymphomas and thymic tumors, appear to be less at risk for perioperative complications than children.85,86

By nature of their anatomic location, these lesions may produce compression of the heart, compression of the large vessels, primarily the vena cava, and compression of the trachea and main bronchi. Anterior mediastinal tumors that are undiagnosed or underestimated as to degree of airway obstruction may completely block the airway on induction of anesthesia and induced muscle relaxation.82,87 Evaluation focuses on an estimate of the presence and degree of obstruction of the tracheobronchial tree and the possibility of avoiding general anesthesia if possible.

Evaluation to assess the patency of the airway at the tracheal and the bronchial level is necessary to formulate an anesthetic plan. By history, symptoms of airway obstruction including dyspnea at rest, on exertion, and in different positions require additional evaluation. The presence of stridor, wheezing, rhonchi, and diminished breath sounds should be reviewed with the patient in different positions. Careful analysis of chest radiographs, CT, and MRI studies may prove essential for planning airway control in the patient with a mediastinal mass. Chest radiographs in the posteroanterior position allow measurement of the tracheal diameter at the level of the clavicles.88 A lateral chest view shows the degree of compression of the trachea in an anteroposterior position. A CT scan of the chest permits accurate measurements of airway diameters and indicates the exact level and extent of compression of the tracheobronchial tree.

Pericardial effusion on preoperative CT scan was the only variable associated with intraoperative complications in a review of 98 patients with mediastinal mass.85 In this population, postoperative respiratory complications were related to tracheal compression of greater than 50% on a preoperative CT scan and a finding of mixed restrictive and obstruction disease on pulmonary function testing. A review of 29 pediatric patients with mediastinal masses who were undergoing general anesthesia concluded that CT evidence of superior vena compression along with symptoms and signs of superior vena cava syndrome (SVCS) were associated with potential development of life-threatening situations.86 In this review, SVCS was the only nonrespiratory sign or symptom that was associated with increased anesthetic risk. All 4 children with SVCS developed acute airway compromise with general anesthesia.

Pulmonary flow volume loop studies performed in the upright and supine positions may sometimes assist in defining the severity of position-related airway compromise. Maximal inspiratory and expiratory flow volume curves may help to quantify the degree of impairment and differentiate extrathoracic from intrathoracic obstruction.66,82 In an evaluation of 37 patients with anterior mediastinal masses by Hnatiuk and Corcoran, however, the incidence of perioperative surgical complications was found to be low. The results of upright and supine spirometry did not always alter the anesthetic technique, and normal spirometry results did not exclude the occurrence perioperative complications.89 In a review of 77 mediastinal mass in patients who underwent pulmonary function tests prior to general anesthesia, airway collapse did not occur in any patient.85 This patient population included 10 cases with a peak expiratory flow rate (PEFR) that was less than 50% of the predicted rate and 6 cases with a PEFR that was 40% or less of the predicted rate. A PEFR of 40% or less of predicted, however, was associated with a more than 10-fold increase in the risk of postoperative respiratory complications.

In general, patients with mediastinal masses are considered at high risk for perioperative complications if they have cardiorespiratory signs and symptoms, tracheal compression more than 50%, pericardial effusion on CT scan, or combined obstructive and restrictive patterns on pulmonary function testing.85 General anesthesia and the use of neuromuscular blocking agents may reduce lung volume, relax bronchial smooth muscle leading to greater compressibility of the airway from the overlying mass, and reduce the transmural pressure gradient across the airway that helps maintain airway diameter.82 A conservative management strategy may be necessary in such patients because mask ventilation may not be possible. Tracheostomy may not relieve airway obstruction because the obstruction may occur at or below the level of the carina. Some clinicians have advocated femoral vessel cannulation in high-risk patients so that cardiopulmonary bypass can immediately be initiated in a crisis.87

Foreign Body

Foreign bodies in the upper aerodigestive tract are an important cause of morbidity and mortality for patients at both age extremes. Both the elderly and children younger than 3 years are at risk for foreign body ingestion. Impacted food in the upper aerodigestive tract tends to be a problem in the elderly. These patients may have dentures that prevent the detection of a small bone fragments or proper mastication of food. In addition, elderly patients are more likely to suffer from esophageal dysmotility, Zenker diverticulum, malignancy, or stricture, all of which predispose to esophageal foreign bodies. Symptoms of an impacted bone include stabbing pain on swallowing. Typically the bone protrudes from the lingual or palatine tonsil. Generally, when a foreign body is lodged in the upper esophagus, the patient can point to the level of obstruction. Dysphagia, regurgitation of food, bloody secretions, and an inability to tolerate secretions may be noted. Removal of an impacted foreign body may be performed under general anesthesia depending on the patient's age, material that is impacted, and location. In these situations, it is imperative to control the airway to prevent aspiration. Even if the foreign body is noted on laryngoscopy, unless it is obstructing the airway, tracheal intubation should be achieved before removal.

Children, particularly those younger than 3 years, can also present with an upper airway foreign body in addition to the more common esophageal foreign body. Airway foreign bodies may involve the larynx, trachea, or bronchi. Most inhaled foreign bodies enter the right mainstem bronchus, which is larger and has a straighter takeoff from the carina than the left. The symptoms associated with aspiration can include gagging, coughing, spasmodic choking, stridor, wheezing, tachypnea, tachycardia, and decreased breath sounds on auscultation.90 Removal of an inhaled foreign body may involve a general anesthetic depending on the patient's age, as well as the location and material of the foreign body. Careful control of the airway in these situations is imperative. The anesthetist should work closely with the otolaryngologist to perform laryngoscopy and bronchoscopy for removal of the foreign body.

The ASA's Closed Claims Project evaluated adverse anesthetic outcomes obtained from the closed claim files of 35 US liability insurance companies. This database dates from 1985 and accrues about 300 cases per year. One of the first reviews of this data evaluated respiratory events, the most common cause of adverse outcomes. This study found respiratory events to be the single largest class of injury accounting for 34% of all claims, with 85% of these adverse outcomes resulting in death or brain damage. Critical review found that most outcomes could have been prevented. It is not surprising that 30% of the mortalities in these claims were attributable to an anesthetic malpractice and were the result of an inability to manage a difficult airway. More recent examination of the data looked at outcomes from perioperative airway claims from 1985 to 1999. In this series, 57% of claims resulted in brain damage or loss of life with the difficult airway being encountered upon induction. Remarkably, in more than half of the claims the difficult airway was anticipated but still was inadequately managed, and 25% of these patients were patients with head and neck pathology.

In an effort to improve management of the difficult airway, the ASA released an airway algorithm in 1993.91 The current edition is only slightly revised and includes the use of supraglottic devices such as the laryngeal mask airway. Since its release, the closed claims data reports a 35% decrease in induction-related airway complications resulting in death and brain damage.1 The study highlights the importance of having in place a plan for airway management as well as the necessary equipment.

A patient with a history of difficult intubation or with conditions associated with difficult airway management should be approached with organized primary and secondary plans for airway management. Often awake intubation, especially if a history of difficult mask intubation is present, is the optimal approach. Patients with physical findings suggestive of a high possibility of difficult intubation may benefit from an awake intubation or the maintenance of spontaneous ventilation during induction of anesthesia. The patient who unexpectedly presents with a difficult or failed intubation history is at most risk because the anesthesiologist may not be prepared.

The hidden cause of a difficult airway (eg, an asymptomatic supraepiglottic cyst) may not be detectable without direct or indirect laryngoscopy. Physicians involved with airway management should be aware of the weaknesses of various tests used to predict difficult airway management and must be prepared to manage an unexpected difficult airway. The ASA's difficult airway algorithm suggests limiting attempts at intubation to avoid trauma. It encourages changing to other techniques to establish ventilation and oxygenation. In a prospective study of 11257 intubations, this predefined algorithm was effective in solving most problems.92 Impossible ventilation never occurred during the 18-month study; 100 cases (0.9%) of unexpected difficult intubation were recorded. A preanesthetic diagnosis of difficult intubation remains a task of recognizing subtle signs, with a tendency for the experienced practitioner to err on the side of a conservative diagnosis because of the difficulty associated with managing an unexpected airway problem.

• American Society of Anesthesiologists Task Force on Management of the Difficult Airway. Practice guidelines for management of the difficult airway: an updated report by the American Society of Anesthesiologist's Task Force on Management of the Difficult Airway. Anesthesiology. 2003;98:1269-1277.
• Combes X, Le Roux B, Suen P, et al. Unanticipated difficult airway in anesthetized patients: prospective validation of a management algorithm. Anesthesiology. 2004;100:1146-1150.
• Goh MH, Liu XY, Goh YS. Anterior mediastinal masses: an anaesthetic challenge. Anaesthesia. 1999;54:670-674.
• Hastings RH, Marks JD. Airway management for trauma patients with potential cervical spine injuries. Anesth Analg. 1991;73: 471-482.
• Henderson JJ, Popat MT, Latto IP, Pearce AC; Difficult Airway Society. Difficult Airway Society guidelines for management of the unanticipated difficult intubation. Anaesthesia. 2004;59:675-694.
• Lee A, Fan LT, Gin T, Karmakar MK, Ngan Kee WD. A systematic review (meta-analysis) of the accuracy of the Mallampati tests to predict the difficult airway. Anesth Analg. 2006;102:1867-1878.
• Rosenblatt WH. The airway approach algorithm: a decision tree for organizing preoperative airway information. J Clin Anesth. 2004;16:312-316.
• Rosenstock C, Gillesberg I, Gätke MR, Levin D, Kristensen MS, Rasmussen LS. Inter-observer agreement of tests used for prediction of difficult laryngoscopy/tracheal intubation. Acta Anaesthesiol Scand. 2005;49:1057-1062.
• Rosenstock C, Hansen EG, Kristensen MS, Rasmussen LS, Skak C, Østergaard D. Qualitative analysis of unanticipated difficult airway management. Acta Anaesthesiol Scand. 2006;50:290-297.
• Shiga T, Wajima Z, Inoue T, Sakamoto A. Predicting difficult intubation in apparently normal patients: a meta-analysis of bedside screening test performance. Anesthesiology. 2005;103:429-437.