Share this article on:

Predicting the Difficult Airway

Braude, Darren MD, EMT-P

doi: 10.1097/01.EEM.0000264923.41762.c6
The Emergency Airway

Dr. Braude is an assistant professor of emergency medicine at the University of New Mexico School of Medicine, the medical director of PHI Air Medical of New Mexico, both in Albuquerque, a co-director of the Emergency Airway Training Program at the University of New Mexico, and a co-director of the National Procedural Sedation Course (



During a shift in a community hospital, an obese 60-year-old trauma patient arrives with an apparent closed head injury. His mental status and oxygen saturation are slowly deteriorating, and you elect to intubate before sending him to CT. Paramedic students ask if they might perform the intubation, saying they have previously intubated five elective medical patients in the OR. How do you quickly assess the patient to determine if he is likely to have a difficult airway or would be a good candidate for these students' first emergency intubation?

What is a difficult airway? I use two definitions: an airway that is anticipated to be difficult and an airway that proves to be difficult (more than one attempt required). Importantly, this refers to difficult intubation and difficult bag-valve-mask ventilation (BVMV). It is worth noting that many OR studies use a much higher standard, such as three or more attempts or an inability to intubate at all.

There are few data on the frequency of difficult intubation in the ED though it appears that about one in 100 intubation attempts fail completely. (Ann Emerg Med 1998;31[3]:325; J Emerg Med 2002;23[2]:131.) In the OR about 10 percent of patients have a “poor view” and two percent of patients require three or more attempts at laryngoscopy while one in 100 to one in 300 intubation attempts fail completely. (Can J Anaesth 1994;41:372; Can J Anaesth 1996;43[1]:30.) Other anesthesia studies have found that 1.5 percent to five percent of patients are difficult or impossible to ventilate using a BVM. (Anesthesiology 2006;105:885; Anesthesiology 2000;92:1229.) It is likely that this number is higher in the ED, though these data are not available.

It would be useful if we could reliably predict which airways are likely to cause us difficulty. For those likely to be difficult, we could call for help in advance, consider alternatives such as awake intubation, and be better prepared by, for instance, having a back-up airway, sized, out of the package, lubricated, and sacrificed to the procedure. In the case of procedural sedation, we may opt for a lower level of sedation or to send the patient to the operating suite.

Common predictors of difficult BVM ventilation known to most emergency physicians include facial trauma, facial hair, obesity, and lack of teeth (assuming you don't have the dentures to replace during BVMV). Other risk factors for difficult BVMV demonstrated in the anesthesia literature include age over 55, history of snoring, Mallampati grade 3 or 4, severely limited jaw protrusion, and thyromental distance less than 6 cm. (Anesthesiology 2006;105:885; Anesthesiology 2000;92:1229.)

Commonly used predictors of difficult laryngoscopy in anesthesia include facial trauma/anomalies, Mallampati grade, thyromental distance, sternomental distance, mouth opening, neck mobility, obesity, and buck teeth. In emergency medicine, some of these features have been assembled into the LEMON mnemonic or 4 Ds. (See box.) (Manual of Emergency Airway Management, 2nd Edition. Lippincott Williams & Wilkins. Philadelphia, 2004; The AirwayCam Guide to Intubation and Practical Emergency Airway Management. AirwayCam Technologies, Inc. Wayne, Pennsylvania 2004.) Unfortunately, most of these clinical assessments cannot be performed in typical ED patients undergoing emergency airway procedures due to their inability to cooperate or sit up. (Ann Emerg Med 2004;44;307.)

The Mallampati score, for example, relies on having a cooperative patient sit up, open his mouth fully, and stick out his tongue so that the extent to which the hard palate, uvula, and posterior pharynx can be visualized can be graded on a 1 (optimal) to 4 (poorest) score. Even at its best, the Mallampati score and other clinical measurements are only modestly predictive of difficulty. (Anesthesiology 2005;103:429.) Combinations of variables may improve their predictive value. (Eur J Anaesthesiol 2003;20[1]:31; Anesth Analg 2006;102:818; Can J Anesth 2000;47[8]:730.)

Some authors also have described a class “zero” Mallampati score in which the tip of the epiglottis is visible on pharyngeal inspection. (Anesth Analg 2001;93[4]:1073.) In an ironic poetic twist, no sooner had these authors reported on the class “zero” airway than other authors reported a patient with a class “zero” airway that could not be intubated! (Anesth Analg 2003;96[3]:911.)

While many of these assessments are neither perfect nor practical on most ED patients, the emergency practitioner should look at gross facial morphology (we are really looking for extremes of disproportion) and in the mouth. I know of two cases where the physician discovered only as he was about to paralyze that his patient had his jaw wired shut!

There are conflicting results about the impact of obesity at laryngoscopy. (Anesthesiology 2005;103:429; Anesth Analg 2002;94:732; Anesth Analg 2003;97[2]:595.) Much of the problem with intubation in the morbidly obese may be overcome with proper positioning, i.e., the “ramped” position. (Obesity Surgery 2004;14:1171.) Obesity definitely makes BVMV more difficult. In addition, common rescue devices such as the LMA Unique may not generate enough airway pressure to lift a very heavy chest. Obesity also limits the effects of pre-oxygenation due to reduced functional residual capacity as well as increased oxygen demand so that time to perform the intubation before critical hypoxemia may be limited.

Another major factor in intubation difficulty is time, a factor ignored in most common clinical scores. Many intubations that are difficult when a patient's saturation is plummeting might be quite doable if time were unlimited. The time-factor in airway management is usually due to oxygen reserve. We have modified the common LEMON mnemonic, adding an “S” for saturations, to make it LEMONS, to take this factor into account. (Ann Emerg Med 2006;47[6]:581.)

Predicting difficult intubation is an imperfect science at best, with limited applicability to most ED patients undergoing airway management. I find it most useful for risk-stratifying cooperative patients I am evaluating for moderate or deep procedural sedation. Prediction of difficult BVMV is somewhat more reliable. Nonetheless, emergency physicians should look for and heed obvious warning signs for a difficult intubation or BVMV (using LEMONS or the 4 Ds) and prepare accordingly. Do not, however, let the absence of any predicted difficulties create a sense of complacency. Any patient, no matter how favorable his airway appears, may prove difficult or impossible to intubate. If you have not encountered such a patient, you have not intubated enough!

In the case that started this column, the patient's obesity, age over 55, limited neck mobility (cervical spine precautions), and limited oxygen reserve are reasons to suspect he will be difficult to intubate and oxygenate with a BVM. Further assessment should at least involve inspection of the mouth and looking for gross anatomical distortion of the face and neck. I would make this a teaching case and involve the paramedic students by having them assist with cricoid pressure, external laryngeal manipulation, or maintaining spinal precautions with or without a jaw thrust, but I would not recommend they attempt this intubation.

Back to Top | Article Outline

Mnemonics for Defining a Difficult Airway


  • L ook externally
  • E valuate the 3-3-3 rule
  • M allampati score
  • O bstruction
  • N eck mobility
  • S aturations

4 Ds

  • D istortion
  • D isproportion
  • D ysmobility
  • D entition
© 2007 Lippincott Williams & Wilkins, Inc.