Approach to the failed intubation

A plan of management for a unexpectedly difficult (i.e. repeatedly failed) intubation is expected of senior ICU trainees. This plan of management inevitably commences with an early "call for help" stage and concludes with some sort of surgical airway, most commonly a cricothyroidotomy.

In the cases considered here, the airway was not expected to be difficult. Recently published data from Nørskov et al (2015) revealed that 93% of the reported cases of difficult intubation were unexpected and not prepared for. Therefore, the algorithms usually begin with direct laryngoscopy.

The scenario has only come up once in the written exam, in  Question 2 from the first paper of 2000. "You are unable to visualise the cords during laryngoscopy. What is your plan to manage this problem?" Subsequent incarnations of this SAQ have been as viva stations (for instance, OSCE 10 from the first paper of 2001, and Viva 8 from the second paper of 2009). This form of assessment probably makes more sense, as a failed airway is as much an exercise in managing stress and tension as it is a cognitive task.

Specific issues to consider in a failed intubation

  • Senior staff presence is desirable; early phone calls to ICU or anaesthetic seniors are strongly suggested. As an example, The Alfred Failed Intubation Plan starts with a call for help.
  • Oxygenation is the priority: to secure this, one should not continue with unskilled attempts at intubation. Ongoing courageous stabbing at the larynx will result in trauma and frustrate the intubation attempts made by others. Instead, focus on bag-mask ventilation.
  • Intubating conditions need to be re-optimised between attempts in order for them to meet with success. This occasionally means that one may merely need to wait for the muscle relaxant to work properly. Many times one witnesses the ICU specialist take over an airway taks, successful only because they were the second person to have a go, and by the time it was their turn the patient had become completely relaxed.

Algorithms for an unexpectedly difficult airway

LITFL has an excellent Difficult Airway Algorithm page, full of algorithms. There are beautiful diagrams available to describe the steps, typically using an eye-hurting combination of primary colours. True to the spirit of bleak nihilism which typifies Deranged Physiology, these colourful offerings have been reinterpreted here as colourless text. Some of these are duplicates from the chapter on preparing for a difficult intubation. One might make the argument that all ICU airways should be considered difficult by default, and prepared for in that fashion.

Difficult Airway Society guidelines (unexpected difficult airway)

  • If the intubation fails, try to insert an LMA or something similar, eg. a Guedel airway
  • If you still can’t bag-mask ventilate, try to face mask ventilate one last time.
  • If you still fail to face-mask ventilate, it’s a CICO and you should go ahead with cricothyroidotomy.
  • The disadvantage of this algorithm is that successes at any stage all seem to lead to waking the patient up, as if this was always an option. The DAS seems content to abort the process of intubation as soon as it is safe to do so, and to give up on the whole idea.

ASA algorithm (Unexpected difficult airway)

  • If you can’t intubate but bag-mask ventilation is easy for you, … keep trying.
    • Those extra attempts should use some sort of “alternative” method, eg. trying different blades or asking another specialist to have a go.
  • If you can’t bag-mask ventilate, you should try an LMA.
  • If the LMA does not work for you, take it out and try to bag-mask ventilate one last time.
  • If neither the LMA nor the face mask are working, it’s a CICO and you need to cut the throat.
  • Again, this is an anaesthesia-oriented algorithm, and most of the technical failures are met with the suggestion “let’s just wake the patient up and cancel the case”. In ICU this approach is patently absurd. You decided to intubate the patient because they are dying of some pathological process which – by your own determination – requires intubation in spite of the fact that it is perceived as difficult, and in the face of all the risks this entails. Clearly, to remain without a tube is not an option here.

ACMEAM (Wall and Murphy, 2008) algorithm (applicable to the unexpected difficult airway)

  • The Manual of Emergency Airway management separates the algorithm into several. There is an Emergency Airway Algorithm, a Crash Airway Algorithm, a Difficult Airway Algorithm and a Failed Airway Algorithm.
  •  Essentially, all of them have three major stages.
  • First, direct laryngoscopy is attempted (perhaps even a few times) while oxygenation can be maintained.
  • If oxygenation can be maintained and there is still no success with intubation, the operator is invited to select among a host of options, which include fiberoptic intubation, videolaryngoscopy, or the use of a lighted stylet.
  • If oxygenation is impossible, a cricothyroidotomy needs to be  performed.
  • This is a reasonable algorithm (or series of algorithms), and it resembles the others insofar as a failure of bag-mask ventilation and LMA oxygenation is treated as an indication for cricothyroidotomy. If the LMA fails to produce an effect, the operator is not invited to pull it out and try the face mask again (as the ASA seem to do) – sensibly, Walls et al advise you to go straight for the throat.

Austin algorithm from the Austin Hospital ED (Melbourne)

  • This is available from LITFL , courtesy of George Douros.
  • Plan A is to try direct laryngoscopy (a maximum of two attempts is permitted)
  • Plan B is to try video laryngoscopy (while calling anaesthetics)- again, two tries.
  • Plan C is to reoxygenate with LMA while waiting for anaesthetics
  • Plan D is a CICO, scalpel-finger-tube.
  • The Austin algorithm is admirable in its prescriptive exactness, including the locally relevant pager numbers and emergency contact phones.  Specific indications for progression through the stages are suggested; for instance Plan D is implemented if the oxygen saturation drops to below 75%. Thankfully, the “wake the patient” option is conspicuous by its absence. Inexplicably, videolaryngoscopy is demoted to the second tier. Some might argue that the first attempt should be one which sacrifices the manly elegance of unaided direct laryngoscopy to maximise success for the patient’s benefit.

CICM (Lim et al, 2003) algorithm

  • Our own college published this thing in Critical Care and Resuscitation
  • It hinges on the suggestion that the first attempt should be a “best laryngoscopy” attempt, involving several simultaneous adjustments to the technique, including:
    • Longer laryngoscope blade
    • BURP
    • A bougie
    • Using a smaller ETT
  • If this fails, and an “alternate airway device” is ineffective, a needle cricothyroidotomy is used to oxygenate the patient while other techniques are being organised.
  • The college offers “alternative intubating technique” as the solution, without being particularly specific about what that means. This is suggested alongside percutaneous and surgical tracheostomy as the solution to a failed intubation with a rapidly decreasing oxygen saturation in spite of a needle in the cricothyroid membrane. The casual onlooker may suggest that this needle might now be used to introduce a guidewire of a Melker emergency cricothyrotomy kit, so that the already established hole may be dilated and converted into a 5-6mm cricothyroid airway. Though the Griggs tracheostomy technique can apparently be performed in around 30 seconds, the only operators who can confidently do this seem to be the people who developed this technique.
  • Another  criticism of this approach is the redundant commentary about “best intubation”, as it implies that the intubation has failed because the first attempt was somehow haphazard and “not best”. Sure, it was probably a crap attempt, but a failed intubation algorithm should not imply that the same operator can court triumph by simply upping their game.  Your first attempt is usually your best attempt.

J.Farkas Respiratory death algorithm (PEA arrest due to hypoxia)

  • This is an algorithm which ensures that the absolute minimum time is wasted on fiddling around with laryngoscopy when it is patently impossible.
  • Double setup: one operator performs laryngoscopy, the other performs a concurrent cricothyroidotomy.
  • Attempts are initialised simultaneously.
  • It’s a race for the airway. The loser aborts their attempt.
  • The disadvantage of this algorithm is fact that the aggression it requires can only be justified in the context of a cardiac arrest due to hypoxia. The patient is already dead; the likelihood of a good outcome depends on how early you can establish oxygenation.  In almost any other circumstances one can easily justify a more measured approach. Additionally, one can envision a nightmarish scenario where poor communication between operators leads to the cricothyroid membrane being punctured just in time for the scalpel to skewer the cuff of a downward-bound ETT.

"Own approach":

  • Change operators: its nothing personal, but a second set of eyes and hands may be the exact thing that is needed.
  • Add airway accessories: insert a Guedel airway and/or a nasopharyngeal airway (or two) and try to bag-mask ventilate.
  • If one is able to ventilate (even using a two-person technique) one can become substantially more relaxed. There is plenty of time to rescue the situation with appropriate equitment. Call for a vidolaryngoscope or intubating LMA.
  • If one is not able to ventilate, one is somewhat screwed. An LMA is essentially the only recourse here. One will need to insert and LMA and try to oxygenate though it.
  • If one is not able to oxygenate via LMA, one will need to go ahead with some sort of surgical option. One may either go straight for a cricothyroidotomy, or to start with a 14G airway cannula for jet insufflation. This should not be the first time one has used the jet ventilation kit.


Nørskov, Anders Kehlet, 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." Anaesthesia 70.3 (2015): 272-281.

Apfelbaum, Jeffrey L., 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." The Journal of the American Society of Anesthesiologists 118.2 (2013): 251-270.

Frerk, C., et al. "Difficult Airway Society 2015 guidelines for management of unanticipated difficult intubation in adults." British journal of anaesthesia 115.6 (2015): 827-848.

Heidegger, T. "The 2015 Difficult Airway Society guidelines: what about the anticipated difficult airway." Anaesthesia 71 (2016): 592-3.

Law, J. Adam, et al. "The difficult airway with recommendations for management–part 2–the anticipated difficult airway." Canadian Journal of Anesthesia/Journal canadien d'anesthésie 60.11 (2013): 1119-1138.

Walls, Ron M., and Michael Francis Murphy, eds. Manual of emergency airway management. Lippincott Williams & Wilkins, 2008.

Lim, M. S., and J. J. Hunt-Smith. "Difficult airway management in the intensive care unit: Practical guidelines." (2003): 43.

Caldiroli, D., and P. Cortellazzi. "A new difficult airway management algorithm based upon the El Ganzouri Risk Index and GlideScope® videolaryngoscope: a new look for intubation." Minerva Anestesiol 77.10 (2011): 1011-1017.

El-Ganzouri, Abdel Raouf, et al. "Preoperative airway assessment: predictive value of a multivariate risk index." Anesthesia & Analgesia 82.6 (1996): 1197-1204.

Zaouter, C., J. Calderon, and T. M. Hemmerling. "Videolaryngoscopy as a new standard of care." British journal of anaesthesia 114.2 (2015): 181-183.

Reynolds, Stuart F., and John Heffner. "Airway management of the critically ill patient: rapid-sequence intubation." CHEST Journal 127.4 (2005): 1397-1412.