Competent airway management is essential in the performance of safe anesthesia. All anesthetics should be preceded by a focused history of previous airway management and physical examination of the airway.1 The objective of airway assessment is not to determine the presence or absence of various clinical signs; instead, the information gained should be used to identify and stratify the risks involved in managing the patient’s airway.
A risk assessment can guide the airway management plan, allowing specialized skills and equipment to be deployed at the bedside. We will show how the airway evaluation can be used to classify patients into three major risk categories, thereby guiding the overall approach to airway planning and management. For the purposes of this article, we will limit the scope of discussion to patients undergoing elective surgery without airway pathology.
Risk assessment is already practiced in many fields, including financial investment, environmental science, and occupational health and safety. Varying definitions of risk assessment exist for each specific field but generally risk assessment involves identifying hazards, quantifying the likelihood of occurrence, evaluating the degree of harm that the threat poses, and preventative actions that can be employed to mitigate the risk. In order to conduct an airway risk assessment, one needs to identify those risk factors that may cause standard airway techniques to be difficult or fail, and to evaluate the likelihood of that difficulty.
The difficult airway is often loosely defined and sometimes poorly understood. Complicating the difficulty can be anatomic, procedural, or contextual factors. For example, difficulty may be experienced during the care of a patient with a physically normal airway but the procedural competence is compromised by a heightened sense of urgency (eg, during endotracheal intubation for an emergency cesarean delivery), leading to a deterioration in human procedural performance. This decline in human performance may be predisposed by poor equipment provision and compounded by limited training or expertise. Procedural difficulty and failure can occur during any of the techniques used to provide airway support.
We will present the existing evidence for increased risk for failure for each aspect of airway management. We will highlight the particular clinical relevance of the predictors of combined difficult bag mask ventilation (BMV) with difficult endotracheal intubation. We will then present a solution to guide decision-making, procedural choice, skill acquisition, and equipment provision at the point of care.
Difficult Bag Mask Ventilation
Much of the existing work to delineate risk factors for encountering difficulty during airway management has focused on the predication of difficult direct laryngoscopy (DL) or endotracheal intubation rather than the ability to perform BMV. Despite its importance in overall airway management and its specific role as the first rescue maneuver after failed endotracheal intubation,1 the ability to perform adequate BMV has only been a focus of investigation relatively recently.
The first major investigation was published in 2000 by Langeron et al, which found a 5% incidence of difficult BMV in a prospective study of 1,502 patients. The outcome was rather loosely defined as “clinically relevant difficulty with BMV which could have led to potential problems if BMV had been maintained for a longer time.”2 This study found that the presence of a beard, body mass index (BMI) greater than 26 kg/m2, a lack of teeth and a history of snoring predicted difficult BMV, according to the definition used (Table).2
The outcome was further studied by Kheterpal et al,3 where difficult BMV was defined as that which is inadequate to maintain oxygenation, unstable, or requires two or more providers. This group reconfirmed the risk factors for difficulty as including the presence of snoring, obesity (BMI >30 kg/m2), and a beard, but also added the modified Mallampati scores III or IV, age older than 57 years, and severely limited jaw protrusion to the list of independent predictors.
Risk factors for impossible BMV received additional scrutiny, the outcome defined as “the absence of end-tidal carbon dioxide measurement and absent chest wall movement during positive-pressure ventilation attempts despite the use of airway adjuvants and two person ventilation attempts.” The specific predictors of impossible BMV were found to be a reported history of snoring and a thyromental distance of less than 6 cm.3In an additional review of over 50,000 cases, the incidence of difficult and impossible BMV (defined as in the previous study) was found to be 2.2% and 0.15%, respectively.4 Independent risk factors for impossible BMV were Mallampati classification III or IV, male sex, presence of a beard, and a history of sleep apnea. However, interestingly, the strongest predictor was the presence of neck radiation changes (adjusted hazard ratio, 7.1), with the authors advising special consideration of the performance of awake intubation in patients with neck radiation changes who have additional risk factors for impossible BMV.
Difficult Supraglottic Airway Placement
The failure rate of different supraglottic airways (SGA) has been variably expressed as 0.2%5 to 4.7%,6 the variance relating to the variety of definitions of failure and type of SGA. Ramachandran et al, in their retrospective review of 15,795 laryngeal mask airway (LMA) insertions performed in a large academic medical center, discovered that 1.1% of patients with an LMA Unique experienced an airway event requiring endotracheal intubation.7 In their study population, they found the independent risk factors for LMA failure to be rotation of the surgical table, male sex, poor dentition, and higher BMI.
The American Society of Anesthesiologists’ (ASA) difficult airway guidelines of 20131 continue to place early SGA use in the algorithm for failed DL with difficult BMV airway rescue, and it is reassuring to note the high success rate of SGA use as a rescue in this event. A description of a case series reported that placement of an SGA restored ventilation in 16 of 17 cases of difficult BMV combined with difficult intubation.8
Difficult Laryngoscopy and Difficult Intubation
Difficult intubation is an often poorly defined outcome, leading to a variety of definitions in clinical use. This variability has hampered investigations into its prediction. A study by Rose and Cohen of 3,325 consecutive adult patients scheduled to undergo DL and intubation (excluding patients anticipated to be difficult undergoing alternative techniques) illustrates this well. They found an extremely wide range of reported intubation difficulty or failure, which varied with the definition used: grade 3 or 4 Cormack-Lehane view = 10.1%; 3 attempts at DL = 1.9%; 4 attempts = 0.5%; and failure to insert an endotracheal tube using DL = 0.1%.9
Difficult intubation has been variably defined according to access (ability to place a device through the mouth), visualization (of the glottis), and eventual endotracheal tube passage. Difficult laryngoscopy is more correctly applied to failure to perform adequate laryngoscopy because of either failed access or poor glottic visualization.
In a meta-analysis of 35 trials reviewing 50,760 apparently normal patients (excluding those with anatomic abnormalities or presumed difficult airways), Shiga et al showed the overall incidence of difficult intubation (defined by them as the presence of a Cormack-Lehane grade 3 or 4 during DL) to be 5.8%.10 The review emphasized the poor predictive value of the existing individual airway measures. Other predictors reviewed included the Wilson risk score, subjective assessments of sternomental distance, mouth opening, Mallampati classification, and thyromental distance.
The ability to predict difficulty improved when airway measures were used in combination; their findings showed that the use of the Mallampati score combined with thyromental distance is an improved predictor (odds ratio [OR], 3.3; area under the curve [AUC], 0.84).
Rose and Cohen analyzed risk factors in 18,205 patients undergoing DL and intubation (alternative approaches to DL were excluded).6 They defined difficult intubation as three or more attempts at laryngoscopy and found the risk factors for this outcome to be male sex, middle age (40-59 years), and obesity. Preoperatively evaluated airway measures associated with an increased risk for difficult intubation included reduced mouth opening (relative risk [RR], 10.3), reduced thyromental distance (RR, 9.7), Mallampati score III or IV (RR, 4.5), and limited neck extension (RR, 3.2).
The utility of combining the Mallampati evaluation with a measure of reduced thyromental distance has been confirmed by other investigators. In a small study of 244 patients, Frerk11 showed a combined positive predictive value (PPV) of 64.3% versus thyromental distance of 7 cm or less alone (PPV, 18.5%) or Mallampati score III or higher alone (PPV, 17.3%) for predicting grade 3 or higher laryngoscopy or the use of an airway bougie. Iohom et al12 used the same definition of difficult intubation, again in a small prospective study (N=212), and found a PPV of 100% when Mallampati score III or higher was combined with either sternomental distance less than 12.5 cm or thyromental distance less than 6.5 cm, whereas individually the PPV was Mallampati, 27%; thyromental distance, 47%; and sternomental distance, 62%.
In the largest study of the performance of the upper lip bite test (ULBT) to predict a limited glottic view on DL (Cormack-Lehane grade ≥3), Myneni et al13 enrolled nearly 7,000 North American subjects. The inability to bite the upper lip was found to have a very low PPV (17.4%) and sensitivity (9.8%). The authors could not recommend this as a useful single measure in the North American population.
A smaller study (N=380) in a Middle Eastern population by Khan et al has shown greater utility of the ULBT.14 Using the same diagnostic criteria, Khan et al showed superior performance of the ULBT to standard airway measures with regard to specificity and accuracy (sensitivity, 78.9%; specificity, 91.9%; PPV, 33.3%; and accuracy, 91.0%). A combination of ULBT and sternomental distance less than 13.5 cm was found to be most sensitive.
Difficult Video Laryngoscopy
The use of video laryngoscopy has increased considerably over the past decade. Aziz et al performed a review of 2,004 GlideScope (Verathon) intubations, performed in two North American academic medical centers, which revealed the GlideScope to be highly successful in patients both with and without established predictors of difficult DL. In this review, 96% of patients with one or more preoperative predictors of difficult laryngoscopy were successfully intubated when the GlideScope was used as the primary laryngoscopic device.15 A striking finding from this study was the preserved high endotracheal intubation success rate (94%) when the GlideScope was used as a rescue device after failed initial DL.15
The data were further analyzed to identify independent predictors of GlideScope intubation failure. These predictors were found to be the presence of abnormal neck anatomy (scar, radiation or mass: OR, 4.39), thick neck (OR, 3.21), thyromental distance less than 6 cm (OR, 2.53), limited cervical motion (OR, 1.76), and institution (OR, 2.28). The differing rates of failure between the two institutions in this study were considered to be related to experience with the GlideScope. The center with better success rates had a greater duration and frequency of GlideScope use, suggesting improved chances of successful intubation with practice and familiarity.
Combined Failure of Airway Management Techniques
The clinical reality of airway management is that when one procedural aspect fails or is difficult (eg, initial BMV), patient harm is commonly avoided by the timely use and success of other procedural interventions (eg, DL or LMA insertion, etc). When two or more of the common airway management modalities (BMV, DL, video laryngoscopy, SGA) fail, this compounds the overall difficulty, limits the rescue methods available, and may lead to failed oxygenation and patient harm.16 Deterioration in oxygenation during airway management is more likely during the care of patients with high intrinsic oxygen requirements (such as in pediatric or obese patients) or related to a pathologic process (such as acute respiratory distress syndrome or sepsis). Studies investigating these complex airway situations are limited but of particular interest and relevance to patient safety.
The relative clinical scarcity of the combined incidence of difficult BMV with difficult DL (3 or 4 Cormack-Lehane view) was revealed by a review performed by the Multicenter Perioperative Outcomes Group of over 170,000 procedures over four tertiary referral academic medical centers,17 with a 0.40% incidence (698 cases) of difficult or impossible BMV in combination with grade 3 or 4 DL. Independent predictors of this combined outcome included age 46 years or older, BMI of 30 kg/m2 or greater, male sex, Mallampati III or IV, presence of a neck mass or radiation changes, limited thyromental distance, sleep apnea, presence of teeth, beard, thick neck, limited cervical spine mobility, and limited jaw protrusion. The authors developed the risk index classification system (or RICS) based on the number of these independent predictors.
Powers of Prediction
The low failure rates of airway management in large North American academic medical centers are revealed by recent large database studies.3,4,17 Given this low incidence, poor prediction is a mathematical certainty exacerbated by the modest sensitivity and specificity of existing airway measures. Our limited current diagnostic ability to predict difficult airway management is catalogued in a recent cohort of 188,064 Danish patients.18
In routine cases (without history of difficulty or airway pathology requiring an awake technique), 93% of difficult intubations and 94% of episodes of difficult BMV remained unanticipated. Conversely, when airway difficulty was anticipated (but in cases where securing the airway using advanced methods, eg, awake fiber-optic intubation, was not deemed necessary) by the anesthesia provider, difficulty was subsequently only encountered in 25% of intubations and 22% of BMV attempts. When all cases were considered, including those predicted to need advanced airway methods, the ability to predict difficulty rose to 60% of difficult intubations in a population with a background incidence of 2.28%.
The airway evaluation performed well as a strong diagnostic test with high positive likelihood ratios (LR+)— difficult intubation, 65.19; difficult BMV, 43.68— when compared with other commonly used tests: eg, Mini Mental State Exam for dementia, LR+ 2.5; ultrasound for proximal deep-vein thrombosis in symptomatic patients, LR+ 47.5; and ST elevation for myocardial infarction, LR+ 11.2. In other words, the standard airway risk assessment seems to perform well at “ruling in” difficulty in those with severe or obvious abnormality but is less effective in detecting difficulty in “normal-appearing” patients. Unfortunately, this latter group is numerically larger, leading to the greatest difficulty being unanticipated in normal-appearing patients.
Given the poor predictive power of our existing preoperative airway risk assessment, even when difficulty is indicated, clinical inertia may conspire to continue with routine airway management rather than to secure the airway by advanced options (video laryngoscopy or awake techniques). The current clinical reality is that almost half of all patients with a predicted risk for difficult intubation in a Danish study were scheduled for standard induction and DL, and difficulty was actually encountered on 1.86% of these occasions.18 Additionally, the ASA Closed Claims database revealed the primary airway management strategy in 61% of cases with anticipated difficult airway to be intubation after induction of general anesthesia.16
Planning for safe airway management needs to overcome this tendency to carry on regardless.
In an attempt to synthesize multiple aspects of the airway evaluation, El-Ganzouri et al developed a risk assessment tool from the examination of a population of over 10,000 cases.19 A weighted score of 7 variables had a greater PPV and higher sensitivity than the highest-performing individual airway measure—the Mallampati score of III (OR, 8.91). Using this simplified airway risk index (SARI), fewer false positives (ie, special intervention for patients who turn out not to have a difficult laryngoscopy) and fewer false negatives (ie, unanticipated difficult laryngoscopy) were encountered compared with use of the Mallampati score. Unfortunately, this trial was underpowered to identify predictors of difficult BMV, so its clinical utility is centered on anticipating difficult laryngoscopy and predicting the need for alternative equipment to DL.
The DIFFICAIR trial20 will prospectively compare standard clinical airway assessment with the SARI to predict difficult DL, and will also compare the ability of systematic documentation of the risk factors for difficult BMV (identified by Kheterpal et al3,4 ) to predict difficult BMV.21 This study should provide insight into whether these predictors generated from large patient series perform better than the standard clinical airway assessment in predicting airway difficulty.
Optimal Clinical Decision Making Using Airway Risk Assessment
The airway risk assessment should lead a provider to a considered opinion regarding the increased incidence of encountering difficulty during airway management before procedural difficulty is encountered. In this way, the provider is afforded the chance to change the initial method of airway management or improve the level of skills and equipment at the bedside. Our recommended approach is to perform a pre-procedural airway risk assessment followed by stratification into three risk categories: likely easy, maybe difficult, and likely difficult (known difficult) (Figure).
Figure. A pre-procedural airway risk assessment stratifies patients into categories of risk, which can be used to plan airway management.
DL, direct laryngoscopy; LMA, laryngeal mask airway; VL, video laryngoscopy
The “likely easy” group corresponds to patients with few to no risk factors for difficulty where proceeding with general anesthesia and securing the airway with DL is appropriate. Typically, these patients have an SARI of 3 or lower or a RICS score of 5 or lower. We earlier discussed that unexpected airway difficulty is unlikely but still encountered in this group, thereby making the presence of standard airway rescue equipment essential.
Patients with more identified risk factors are considered at an incrementally higher risk for difficulty; however, we have discussed how the true incidence remains in a “clinical gray zone.” Our preparations for airway management must be guided by an educated clinical decision performed on a case-by-case basis, based on the presence or absence of risk factors.
For patients considered “at higher risk for difficulty,” this clinical decision must initiate the consideration of a series of airway management plans, each with skills and equipment to support their performance at the patient’s bedside (for instance, the presence of a video laryngoscope and the skills to use it). Many of these patients will ultimately be easily managed with conventional BMV and DL, but the presence of back-up plans serves as a safety net for the minority.
Ensuring correct equipment provision, optimal head positioning (with the use of the ramped position, if required), and thorough preoxygenation before induction of anesthesia should occur before any airway management procedure. However, this has particular importance after the identification of increased risk to ensure the best chance of success on the initial attempt and continued patient safety during additional attempts. A video laryngoscope could be used during the initial laryngoscopic attempt or in the event of failed DL. A correctly sized SGA should be available for use in the event of failed BMV, or to facilitate a conduit approach to endotracheal intubation.
As discussed, the presence of overlapping risk factors predicting difficulty and failure of both difficult DL and BMV4 affords this group particular risk. When considering the existing scoring systems, those subjects in the “at higher risk” group have a SARI of 4 or higher or a RICS score of 6 or higher.
Subjects with a known difficult airway (eg, identified from information in the electronic health record or Medic Alert bracelet) or who have multiple significant risk factors for failure of BMV, intubation, or video laryngoscopy, or simultaneous failure in several of these management modalities (eg, difficult or impossible BMV with DL) should be considered to be in the “known difficult” group.
In an elective context, a technique that retains spontaneous ventilation should be used. Our preferred strategy would be an awake fiber-optic intubation in those requiring a general anesthetic (awake video laryngoscopy is another alternative). Severely limited mouth opening, combined with limited neck extension (precluding DL, SGA, and video laryngoscopy), should be considered at very high risk for difficulty due to the potential failure of multiple airway management modalities. Additional consideration should be given to the performance of an awake technique in patients who would otherwise be thought to be at “higher risk/maybe difficult” but who have impalpable anatomic landmarks or contraindications for surgical airway access.
The overall clinical context of the patient and medical team may also influence the airway management category. For instance, an awake technique may be considered in a patient only thought to be at higher risk for difficulty but with severe gastroesophageal reflux disease to avoid aspiration during multiple attempts at intubation. Similarly, an awake technique may be considered in a patient with borderline pulmonary physiology preoperatively who may not tolerate the suboptimal ventilation generated by an improperly fitting face mask or SGA. The practitioner’s skill level with different airway apparatuses and the availability of these items also affect which strategy is used.
Unfortunately, our current airway evaluation has limited discriminatory ability due to its poor predictive power and the scarcity of the outcome of true difficulty during airway management. Given this inherent unpredictability, difficulty (unanticipated or otherwise) will continue to occur during the various aspects of airway procedural performance. However, recent investigations have provided us with a selection of risk factors that independently predict airway procedural difficulty and failure. The challenge is in how this diverse information can be used to determine our pre-procedural decision-making.
This review has presented the risk factors associated with procedural failure of the individual components of overall airway management, with particular emphasis on the clinical importance of combined procedural failure. Then we presented a description of how the identification of known risk factors can be used to support clinical decision-making by risk assessment and categorization of patients into those whose airway management is considered to be likely easy, at higher risk for difficulty, and known difficult (likely difficult). This type of decision-making allows us to modify airway management plans and ensure that additional equipment (and skills to use them) are present at the point of care when required.
Additional work is needed to improve the predictive power of those airway measures (used individually or in combination) informing our airway risk assessment. In the meantime, we must improve our clinical application of risk assessment with the help of cognitive aids such as checklists, flowcharts, and real-time airway risk assessment calculators embedded within the electronic health record. Additional technological innovation will improve our procedural success during predicted and unexpected difficult airway management. Until such improvement is realized, we must ensure that our airway evaluation, decision-making, technical skills, and equipment provision are optimized to ensure patient safety.
- Apfelbaum JL, Hagberg CA, Caplan RA, et al. Practice guidelines for management of the difficult airway: an updated report by the American Society of Anesthesiologists Task Force on Management of the Difficult Airway.Anesthesiology. 2013;118(2):251-270.
- Langeron O, Masso E, Huraux C, et al.. Prediction of difficult mask ventilation. 2000;92(5):1229-1236.
- Kheterpal S, Han R, Tremper KK, et al. Incidence and predictors of difficult and impossible mask ventilation. 2006;105(5):885-891.
- Kheterpal S, Martin L, Shanks AM, et al.. Prediction and outcomes of impossible mask ventilation: a review of 50,000 anesthetics. 2009;110(4):891-897.
- Verghese C, Brimacombe JR. Survey of laryngeal mask airway usage in 11,910 patients: safety and efficacy for conventional and nonconventional usage.Anesth Analg. 1996;82(1):129-133.
- Rose DK, Cohen MM. The airway: problems and predictions in 18,500 patients.Can J Anaesth. 1994;41(5 Pt 1):372-383.
- Ramachandran SK, Mathis MR, Tremper KK, et al. Predictors and clinical outcomes from failed Laryngeal Mask Airway Unique™: a study of 15,795 patients. 2012;116(6):1217-1226.
- Parmet JL, Colonna-Romano P, Horrow JC, et al. The laryngeal mask airway reliably provides rescue ventilation in cases of unanticipated difficult tracheal intubation along with difficult mask ventilation.Anesth Analg. 1988;87(3):661-665.
- Rose DK, Cohen MM. The incidence of airway problems depends on the definition used.Can J Anaesth. 1996;43(1):30-34.
- Shiga T, Wajima Z, Inoue T, et al. Predicting difficult intubation in apparently normal patients: a meta-analysis of bedside screening test performance. 2005;103(2):429-437.
- Frerk CM. Predicting difficult intubation. 1991;46(12):1005-1008.
- Iohom G, Ronayne M, Cunningham AJ. Prediction of difficult tracheal intubation.Eur J Anaesthesiol. 2003;20(1):31-36.
- Myneni N, O’Leary AM, Sandison M, et al. Evaluation of the upper lip bite test in predicting difficult laryngoscopy.J Clin Anesth. 2010;22(3):174-178.
- Khan ZH, Mohammadi M, Rasouli MR, et al. The diagnostic value of the upper lip bite test combined with sternomental distance, thyromental distance, and interincisor distance for prediction of easy laryngoscopy and intubation: a prospective study.Anesth Analg. 2009;109(3):822-824.
- Aziz MF, Healy D, Kheterpal S, et al. Routine clinical practice effectiveness of the Glidescope in difficult airway management: an analysis of 2,004 Glidescope intubations, complications, and failures from two institutions. 2011;114(1):34-41.
- Peterson GN, Domino KB, Caplan RA, et al. Management of the difficult airway: a closed claims analysis.Anesthesiology. 2005;103(1):33-39.
- 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. 2013;119(6):1360-1369.
- Nørskov AK, Rosenstock CV, Wetterslev J, et al. Diagnostic accuracy of anaesthesiologists’ prediction of difficult airway management in daily clinical practice: a cohort study of 188 064 patients registered in the Danish Anaesthesia Database. 2015;70(3):272-281.
- el-Ganzouri AR, McCarthy RJ, Tuman KJ, et al. Preoperative airway assessment: predictive value of a multivariate risk index.Anesth Analg. 1996;82(6):1197-1204.
- Nørskov AK, Rosenstock CV, Wetterslev J, et al. Incidence of unanticipated difficult airway using an objective airway score versus a standard clinical airway assessment: the DIFFICAIR trial – trial protocol for a cluster randomized clinical trial. 2013;14:347.
- Nørskov, AK. The Difficult Airway Management Trial: “The DIFFICAIR-Trial”. In: ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine (US). 2000- .