A 56-year-old male was the driver of a motorcycle involved in a motor vehicle crash. The primary survey performed by the trauma team reveals the patient to have a patent airway, with spontaneous breathing and clear, bilateral breath sounds. His circulation is normal with strong pulses present in all limbs and no signs of external hemorrhage. His Glasgow Coma Scale is 14 of 15 due to slight confusion. Secondary survey reveals rib fractures on the left and several orthopedic injuries including a broken tibia and femur on the left as well as a pelvic fracture. Several hours after the injury, the patient is brought to the operating room (OR) for surgical repair of the long bone fractures and external fixation stabilization of the pelvic fracture under general anesthesia.

He is admitted to the ICU immediately after the surgery, and is extubated uneventfully shortly thereafter. For the first 2 days after the injury, the patient has required high doses of opioids to control his pain from the rib fractures. Late in the evening of the third postoperative day, he spikes a temperature of 38.9°C, and is having trouble clearing his secretions with coughing. His respiratory rate is 32 per minute; his oxygen saturation is 92% on a non-rebreather face mask (NRM); his heart rate is 120 beats per minute and his blood pressure is 160/85. He is complaining of dyspnea and severe pain. The patient looks tired and has obvious use of accessory muscles of respiration. A chest radiograph reveals volume loss and airspace disease with air bronchograms in the right lower lobe but no pneumothorax.

The house staff is alerted by the patient’s nurse. A first-year resident is on call with a senior fellow in critical care medicine (CCM). After their assessment and discussion with the attending staff by telephone, the house staff team proceeds with endotracheal intubation. Based on the examination of the airway, no difficulties are anticipated with the tracheal intubation itself. The respiratory therapist prepares the usual equipment while oxygen is administered by NRB. The resident physician administers 2 mg of midazolam and with the patient placed in semi-Fowler position at 45 degrees, makes an attempt at intubation using direct laryngoscopy (DL) under the supervision of the senior fellow. The first attempt results in an esophageal intubation. Oxygen saturations dip into the high 80s. After repositioning the patient’s head and neck, and recovery of the saturations into the low 90s this time employing a bag-mask unit with high-flow oxygen, a second attempt is performed by the resident. Despite using a tracheal introducer (also known as “gum-elastic bougie”) and the BURP maneuver, the attempt is again unsuccessful.

A third attempt at tracheal intubation is made by the CCM fellow following the administration of propofol 100 mg and succinylcholine 120 mg, but this also fails. After this intubation attempt, it becomes more difficult to manually ventilate the patient using bag-mask. It is unclear if this is because neuromuscular function has returned or if it is related to deteriorating lung compliance. The patient needed several boluses of vasopressor to treat hypotension during this last intubation attempt, probably related to high bag-mask-ventilation (BMV) pressures coupled with the circulatory effect of the propofol. Recognizing the seriousness of the situation, the CCM fellow requests that some more airway equipment be made available in the event that they cannot secure the airway. The respiratory therapist reports that some of the difficult airway equipment is down the hall in a locked room, and that he needs to leave to acquire the items requested. None of the staff involved in the resuscitation are aware of what exactly is available on this cart, but clearly, not having it in the room from the beginning was an error in judgment.

The fellow now requests that the anesthesia team be called to assist with securing the airway. A video-laryngoscope is brought from the OR by a senior anesthesia resident and is used to successfully intubate the trachea of the patient. Video-laryngoscopy reveals significant swelling of the supraglottic structures and vocal cords.

Those involved in the care of this patient recognized that this situation could have had a more negative outcome. A multidisciplinary meeting was scheduled to debrief and review options for improving their approach to airway management in the ICU employing a root cause analysis (RCA) methodology and intended to implement systematic changes that are preventative in nature.

Is Airway Management in the ICU Associated with More Complications Compared to the OR?

Airway management outside the controlled OR environment carries higher risk to the patient for a variety of reasons including the acuity of the situation, the patient’s limited physiologic reserve, and less access to advanced airway equipment. In addition, airway management is often performed by practitioners with limited experience, commonly after regular working hours when more experienced help may be harder to summon.¹

Perhaps the most common indication for out of the OR but in hospital airway management is cardiac arrest. It is known that cardiac arrest outside of the OR occurs frequently.² Furthermore, the incidence of cardiac arrest in the ICU setting is as high as 2%, much higher than the 0.068% rate in the OR.³ Chacko et al.⁴ reported on critical incidents in a closed 18-bed, multidisciplinary unit over a 33-month period. Airway-related incidents accounted for 32.8% of all reported incidents, with the most common being accidental extubation. In this study, there were 32 incidents (11.4% of those reported) that led to adverse outcomes, including 4 deaths, all of which were due to airway-related events.

Needham et al.⁵ reported on factors which contributed to airway events that had been collected as part of the Intensive Care Unit Safety and Reporting System, a voluntary anonymous reporting system developed in conjunction with the Society of Critical Care Medicine and used in 18 ICUs across the United States over a 12-month period. There were 841 incidents reported with 78 airway events. More than half of the airway events were considered preventable and about 20% of the patients with airway reports sustained a physical injury and had an actual or anticipated prolonged hospital length of stay associated with the event. There was one death related to an airway event. Additionally, family dissatisfaction was common when these events occurred. Factors noted to limit adverse airway events included adequate ICU staffing and the use of skilled assistants.

The Fourth National Audit Project of the Royal College of Anaesthetists and Difficult Airway Society (NAP4)⁶ reported a significantly higher rate of serious airway complications occurring in the intensive care unit (ICU) compared to the OR. In 61% of the ICU cases reported to NAP4, these complications led to death or persistent neurological injury, in contrast to 33% of cases reported from emergency departments and 14% of the OR cases. In addition, NAP4 also identified that 70% of the cases and 60% of the deaths in ICU involved complications of tracheotomy. In this audit, airway events in the ICU were more likely to occur after hours, and be managed by a clinician with less anesthesia experience. Further review of these cases revealed deficiencies that included poor identification of high-risk patients, poor or incomplete planning, and inadequate provision of skilled staff and equipment to manage the events.

Why Are Airway Interventions in the ICU More Dangerous?

Patients admitted to the ICU generally have limited physiologic reserve. They need minute-to-minute monitoring and treatment and usually require respiratory and/or hemodynamic support. Physicians choosing to practice in the ICU environment must possess excellent airway management skills for a variety of reasons. Some patients initially without tracheal intubation will decompensate while in ICU and require tracheal intubation. Others are admitted to the ICU having been intubated elsewhere and will be extubated (planned or unplanned) during their ICU stay. Still others, having been extubated, will fail and require reintubation. As with any intervention, an understanding of the dynamic and often subtle interplay between commonly utilized medications and the physiologic reserve of the compromised patient must be understood.

A particular risk is the patient presenting to the ICU with “severe dyspnea” which is related to severe metabolic acidosis (e.g., ethylene glycol, methanol, or salicylates). Often, the patient is thought to be anxious. Obviously, the administration of sedative agents or opioids in these patients is a recipe for disaster.

Denitrogenation is more difficult in the critically ill patient population, making oxygen desaturation during airway management a common adverse event. Oxygen desaturation is associated with serious complications such as cardiovascular collapse, global hypoxia related brady-asystolic arrest, dysrhythmias, neurologic injury, or death. There are important physiological reasons for such limited respiratory reserve:

• 
  

  increased oxygen consumption in the critically ill patient leads to reduced safe apnea time and an increased alveolar-arterial gradient makes the saturation of hemoglobin less efficient with a given functional residual capacity⁷;

  
  
• 
  

  this is compounded by an increased shunt fraction in some patients related to diffusion block (e.g., pulmonary edema) and reductions in functional residual capacity (FRC);

  
  
• 
  

  predictable reduction in FRC related to obesity, position (increased closing capacity), lung disease, etc. attenuates safe apnea time even further.

Several strategies may be employed in improving denitrogenation and prolonging the safe apnea period. These include⁸:

• 
  

  positioning the patient in a semi-recumbent position;

  
  
• 
  

  continuous nasal cannulae at 10 to 15 L·min⁻¹;

  
  
• 
  

  noninvasive ventilation if oxygen saturations are 90% or less, titrating PEEP to between 5 and 15 cm H₂O;

  
  
• 
  

  use of a PEEP valve during BMV to augment the patient’s tidal volume.

Preemptive or continuing use of Transnasal Humidified Rapid-Insufflation Ventilatory Exchange (THRIVE)⁹ (e.g., Optiflow^®) should be considered. Simple high-flow nasal oxygenation does not assist with ventilation in apneic patients and therefore arterial carbon dioxide concentrations will rise at a predictable rate. However, the Optiflow© system has been shown to provide some degree of ventilation, as measured by slower rise of carbon dioxide over time.

Denitrogenation should be provided for at least 3 minutes, situation acuity permitting. Delayed sequence intubation (DSI)¹⁰ employing ketamine titration balancing continued patient ventilation with CNS obtundation has been used successfully in critically ill patients to aid the denitrogenation process.

Hypotension at the time of intubation is more common in critically ill patients and is associated with an increased mortality. Risk factors associated with an increase in post-intubation hypotension are septic shock, renal failure, respiratory failure, and increased age.¹¹ Right ventricular failure is an especially difficult condition to manage should it preexist, or result from the acute respiratory event. Several strategies are useful in addressing this important complication: ensuring patent intravenous access, assessing the need for a fluid bolus prior to intubation (in the absence of pulmonary edema), having vasopressor therapy immediately available and in some cases preemptively administered as a bolus at the time of induction or by continuous infusion.

In addition to limited cardiopulmonary reserve, other important concerns, such as hepatic and renal dysfunction, a “full stomach” (e.g., patients on continuous tube feeding), and altered neurological function are also common in this patient population.

It is not uncommon in the ICU population for patients to be extubated and then need to be reintubated at a later time in their care. A prospective database of 1053 ICU intubations revealed that subsequent intubations have an even higher risk of complications compared to the first intubation.¹² This increase in complications occurred without there being a measurable increase in predicted technical difficulty. The main complications were hypotension and hypoxemia. This study highlights the importance of recognizing that these patients are at higher risk of these complications and the need to strive to avoid them with proper preparation.

Intensive care patients are as complex as the environment in which they are cared for. The practitioner tasked with airway management in this setting must be aware of these factors and make sound decisions, often very quickly, to improve patient outcome.

How Can We Improve Outcomes Related to Airway Interventions in the ICU?

The literature clearly identifies the ICU population as a group of patients at higher risk of life-threatening complications related to airway interventions, but how can we improve outcomes for these patients related to airway interventions? Several approaches have been identified to improve the culture of safety and optimize the care of airway emergencies outside the OR. Some institutions have organized a group of multidisciplinary experts to focus on improving safety and quality improvement with respect to airway management, such as one institution’s so-called DART (difficult airway response team).¹³ This approach focuses on safety monitoring, quality improvement, availability of equipment in key locations, an educational program, and a response team that is deployed in an anticipated or unanticipated difficult airway situation.

What Are the Pharmacologic Considerations for Airway Management in the Critically Ill?

Several of the pharmacologic agents routinely used in an elective situation could be harmful to a critically ill patient with a limited reserve. A detailed discussion on the pharmacology of intubation can be found in Chapter 4. Much has been debated about the best pharmacologic approach to facilitate airway management in the critically ill. The important question is how to provide the best conditions for intubation with the least amount of risk. In particular, it is important to mitigate the hemodynamic swings that may result from these agents as well as to consider whether neuromuscular blocking agents should be used. In addition, several factors other than pharmacologic agents may contribute to severe hemodynamic instability in the critically ill patient (e.g., institution of positive pressure ventilation [PPV], especially in the presence of hypovolemia). It is best to prepare for and expect hemodynamic instability associated with airway management in these patients regardless of the agents used by preemptively administering vasopressors by infusion or boluses depending on the situation.

The ideal induction agent should provide good conditions for tracheal intubation without adversely impacting cardiac contractility or systemic vascular resistance. Etomidate has been used as an induction agent in the hemodynamically fragile patient because it is thought to have a favorable hemodynamic response compared to propofol. However, there is growing concern that use of even a single bolus of etomidate is associated with adrenal insufficiency and increased mortality.¹⁴ When completed, the KEEP PACE trial, comparing ketamine with etomidate for inducing critically ill patients may shed some light on this question.¹⁵