The physiologically difficult airway: Optimization of airway management in adults at risk of decompensation for emergency medicine and critical care

INTRODUCTION — 

A physiologically difficult airway is present when the patient's underlying comorbidities (eg, heart failure, lung disease) or acute conditions (eg, sepsis-induced hypotension, outflow obstruction by pulmonary embolism) create a significantly higher than usual risk of cardiopulmonary decompensation during airway management. Determining the presence and severity of characteristics associated with a physiologically difficult airway is of paramount importance. Key concepts and techniques for identifying and reducing the risk associated with the physiologically difficult airway are reviewed here. Other aspects of emergency airway management, including the anatomically difficult airway, are discussed separately. (See "Approach to the anatomically difficult airway in adults for emergency medicine and critical care" and "Rapid sequence intubation in adults for emergency medicine and critical care".)

TERMINOLOGY AND EPIDEMIOLOGY

What is a physiologically difficult airway? — Critically ill patients suffer a high rate of intubation-related morbidity and mortality. Studies in the era of modern airway management report that between 20 and 50 percent of intubations performed in critically ill patients are associated with an adverse event, most commonly hypoxemia, hypotension, or cardiac arrest [1-3].

Historically, the difficult airway was defined as one for which difficulty stemmed from anatomic characteristics of the patient that made intubation, mask ventilation, or rescue oxygenation difficult. As a result, first-pass success (or first-attempt success) became an important goal for emergency tracheal intubation, as it was associated with the greatest reduction in airway management-related harm. Strategies shown to improve first-pass success rates include the use of video laryngoscopy and rapid sequence intubation (RSI) using a sedative and neuromuscular blocking agent in rapid succession. (See "Rapid sequence intubation in adults for emergency medicine and critical care".)

Despite these strategies, a high number of critically ill patients experience complications despite first-attempt success or the absence of anatomic obstacles. This group of patients is at risk because of underlying physiologic abnormalities that increase the risk of rapid oxygen desaturation, cardiovascular collapse, or both regardless of the presence or absence of challenging anatomic characteristics. This is the physiologically difficult airway [1].

Underlying physiologic abnormalities in these critically ill patients can increase their sensitivity to the side effects of induction agents, attenuate or even eliminate the effectiveness of preoxygenation, and exaggerate the adverse effects of cessation of spontaneous breathing and transition to positive pressure ventilation.

How often does physiologic decompensation occur during the peri-intubation period? — Most studies on intubation in critically ill patients report that the rate of serious adverse events ranges from approximately 20 to 30 percent, between one in five and one in three patients [2,3]. The most common complications are oxygen desaturation and hypotension. The rate of peri-intubation cardiac arrest is 2 to 4 percent across all intubations in critically ill patients, with higher odds in patients with hypoxemia or hypotension.

The multinational INTUBE study reported that nearly half of critically ill patients suffered an intubation-related adverse event, most often cardiovascular instability [4]. In a large observational study of intubations performed outside the operating room, patients with difficult airways (>2 intubation attempts) were at significantly higher risk of severe complications [5]. Subsequent research has established the importance of first-pass success, which has been associated consistently with reductions in physiologic decompensation [6]. However, even when the first intubation attempt is successful, there remains a significant risk of serious adverse events among critically ill patients.

Pre-existing physiologic abnormalities are the biggest contributing factor to complications in this group [7]. Among critically ill patients with characteristics of a physiologically difficult airway, multiple observational studies of intubations performed in the emergency department and intensive care unit report that significant risk for serious adverse events persists during the peri-intubation period despite first-pass success [8-10]. As further examples, studies performed during the coronavirus disease 2019 (COVID-19) pandemic reported a significant peri-intubation complication rate, largely from desaturation in patients with hypoxemic respiratory failure, despite an increase in first-pass intubation success [11-13].

Patients at risk for physiologic decompensation during the peri-intubation period

Overall risks — The risk profile of physiologic decompensation differs for oxygen desaturation and cardiovascular collapse. Evidence suggests that in at-risk patients, the incidence and severity of desaturation increase linearly with successive attempts at intubation (thus the importance of first-pass success), while the risk of cardiovascular collapse remains fairly constant through multiple intubation attempts [4,14].

The risk of oxygen desaturation depends on the complex relationship between the underlying disease state and the efficacy of preoxygenation. The risk of cardiovascular collapse depends on the nature of the underlying hemodynamic abnormalities and how these are exacerbated by induction medications and the effects of positive pressure ventilation.

Several studies have explored factors associated with peri-intubation decompensation. A systematic review and meta-analysis of 44 studies, including over 34,000 intubations in critically ill patients, found differences in rates of major adverse events by site (intensive care unit 41 percent versus emergency department 17 percent) and reported correlations between major adverse events, hemodynamic instability before intubation, and induction with propofol [15]. Intubation for airway protection was associated with lower rates than intubation for respiratory failure. In a large, international, prospective study of intubations in critically ill patients, major adverse events during the peri-intubation period were associated with the following [4]:

●Higher median Sequential Organ Failure Assessment (SOFA) score (calculator 1)

●Greater severity of heart failure (table 1)

●Bilateral pulmonary infiltrates or pleural effusions

●Poor oxygenation (based on measurement of PaO₂/FiO₂ or SpO₂/FiO₂)

●Treatment with vasopressors or fluids for hypotension

Cardiac arrest — Based on data from clinical trials and observational studies, the following factors are associated with peri-intubation cardiac arrest [16,17]:

●Preintubation hypotension (eg, systolic blood pressure <90 mmHg despite fluid resuscitation)

●Preintubation hypoxemia (SpO2 nadir <80 percent or decrease in SpO2 >10 percent if maximal preintubation SpO2 <90 percent)

●No preoxygenation performed prior to intubation

●High body mass index (>25 kg/m²)

●Age >75 years

The odds of cardiac arrest increase nonlinearly with the addition of each risk factor, from 1.31 with one risk factor to 9.89 with all five risk factors.

Hypotension (cardiovascular collapse) — In multiple observational studies, commonly identified risk factors for peri-intubation hypotension include:

●Preintubation hypotension

●Preintubation vasopressor use

●Sepsis (distinct risk factor from hypotension)

●Left heart failure

●Right ventricular dysfunction/failure

●Recent diuretic use (<24 hours)

●Increasing age

●Increasing severity of illness

●Propofol used for induction of RSI

●Etomidate used for induction of RSI

●Ketamine used for induction of RSI

In a prospective, multicenter observational study of 934 intensive care unit patients, the following were found to be important risk factors for peri-intubation hypotension: mean arterial pressure <65 mmHg, systolic pressure <130 mmHg, sepsis diagnosis, diuresis in preceding 24 hours, vasopressors immediately prior to intubation, etomidate as the induction agent, increased age, and increasing severity of illness [18,19]. These findings are consistent with those reported in a secondary analysis of the INTUBE study [20]. In this study and a meta-analysis, induction with propofol was identified as an independent risk factor for hypotension [15]. Patients with moderate or severe right ventricular dysfunction are also at increased risk of peri-intubation cardiovascular collapse [21].

Oxygen desaturation — Patients intubated for respiratory failure and those requiring noninvasive respiratory support are at risk for oxygen desaturation (SpO2 ≤88 percent) during intubation [15,22,23]. In a prospective observational study of 1033 critically ill adults requiring intubation, risk factors for peri-intubation hypoxemia included [24]:

●Use of noninvasive ventilation before intubation

●Difficult mask ventilation

●Emergency intubation

●Cardiac reason for intubation

●Fluid resuscitation within the preceding 24 hours

PROBLEM-BASED ASSESSMENT AND MANAGEMENT

Hypoxemia

Causes of ineffective preoxygenation and peri-intubation hypoxia — Preoxygenation is an essential step in rapid sequence intubation (RSI). It provides a longer period before clinically significant oxygen desaturation occurs during the apneic period, regardless of the patient's condition, age, or body habitus. Common conditions that cause oxygen desaturation or exacerbate hypoxemia during RSI and interventions to help prevent or manage them are summarized in the following table (table 2). (See 'Preoxygenation (denitrogenation)' below.)

Three phenomena can contribute to ineffective preoxygenation and increase the risk for acute hypoxemia during the peri-intubation period:

●Inadequate oxygen flow for denitrogenation – Inadequate preoxygenation occurs in patients with a high respiratory effort that dilutes supplemental oxygen with ambient air when delivered by standard methods. This problem can arise in patients with severe tachypnea receiving preoxygenation with a nonocclusive nonrebreather face mask. Clinical examples include patients with pneumonia and high inspiratory effort and patients with sepsis and high compensatory minute ventilation. Management is discussed below. (See 'Inadequate preoxygenation' below.)

●Reduced functional residual capacity (FRC) – FRC is the volume of air in the lungs after normal exhalation. The FRC, normally 30 mL/kg in a healthy adult, provides the reservoir of oxygen during the apneic phase of RSI created by denitrogenation.

FRC is decreased by any source of external compression (eg, obesity, ascites, pleural effusion, pneumothorax, or hemothorax) or by functional parenchymal volume loss (eg, pneumonia, acute respiratory distress syndrome [ARDS], heart failure). The safe apneic period is shorter in proportion to the reduction in FRC. In patients with reduced FRC, increasing end-expiratory lung volume is the priority during preoxygenation. Management is discussed below. (See 'Reduced functional residual capacity' below.)

●Low ventilation to perfusion ratio (V/Q mismatch) – A low ratio of ventilation to perfusion occurs in patients with intracardiac right-to-left shunt or intrapulmonary shunt (eg, arteriovenous malformations, hepatopulmonary syndrome) and those with ARDS and hypoxemia refractory to advanced preoxygenation strategies. Further detail about susceptible patients and management methods is provided below. (See 'Low ventilation-perfusion ratio (high shunt)' below.)

Continuous monitoring of oxyhemoglobin saturation by pulse oximetry or continuous arterial monitoring is essential during RSI. This task should be assigned to an individual who can reliably track and regularly report the information without other tasks or distractions. Pulse oximetry readings obtained with a finger probe may lag behind the central arterial circulation, particularly in critically ill patients, and thus give an incorrectly high oxygen saturation [25,26].

Preoxygenation (denitrogenation) — Preoxygenation is an essential step in RSI. Preoxygenation is reviewed in detail separately; aspects of special relevance to the physiologically difficult airway are discussed below. (See "Rapid sequence intubation in adults for emergency medicine and critical care", section on 'Preoxygenation'.)

●Mechanism of preoxygenation – Ensuring that end-expiratory alveolar gas contains a high concentration of oxygen is necessary to maintain a safe level of oxygen saturation during the apneic period of RSI. Normally, end-expiratory alveolar gas comprises primarily nitrogen with small fractions of water vapor, residual oxygen, and CO₂. "Denitrogenation" involves delivering gas with a high fraction of oxygen (FiO₂) to displace nitrogen, such that the end-expiratory alveolar gas contains mostly oxygen, along with water vapor and residual CO₂.

Performance of preoxygenation (denitrogenation) — Overcoming limitations to preoxygenation in the critically ill can be accomplished by two methods:

●Deliver high concentrations of oxygen at high rates – High flow rates of high FiO₂ oxygen can be achieved with "flush" flow oxygen delivered through a nonrebreathing mask by opening the valve on the oxygen inflow completely. This fills the reservoir with 20 to 80 liters per minute (depending on the hospital's oxygen system pressure) and increases the FiO₂ of the gas around the mask as oxygen escapes the bag through the valve [27,28].

●Ensure a tight mask seal – Air leakage can be prevented by using a tight-fitting face mask, such as a noninvasive positive pressure ventilation (NPPV) mask, or a high-flow nasal oxygen (HFNO) system.

Denitrogenation can be supplemented further by the following means:

●Maintaining high-flow oxygen until the patient is apneic.

●Performing gentle mask ventilation in the interval between induction and laryngoscopy.

●Performing oxygenation during the apneic period by providing high oxygen flows through standard nasal cannula or a high-flow nasal cannula system.

Classic instruction for preoxygenation during RSI is to have the patient breathe 100 percent oxygen (FiO₂ of 1) for three minutes or eight vital capacity breaths. However, this approach has proven inadequate in critically ill patients, and even extending the duration of preoxygenation fails to reduce the risk of desaturation [29]. The strategy fails primarily because of equipment limitations and the patient's inspiratory flow demands.

The nonrebreathing masks commonly used in emergency departments and hospitals have a continuous flow of oxygen (usually 15 liters per minute) into a reservoir (figure 1). When the patient breathes, oxygen is drawn from the reservoir across a one-way valve. Exhaled air exits the mask through another one-way valve, which prevents the patient from "rebreathing" exhaled air. However, because the mask does not have a tight seal, ambient air (FiO₂ 0.21) is entrained and dilutes the oxygen from the reservoir, decreasing the FiO₂ being inspired. Patients with a high inspiratory effort (eg, due to respiratory distress) draw more ambient air than patients with a low inspiratory effort, thus this method is least effective in patients who need it the most.

Many critically ill patients receive preoxygenation via a bag-valve mask only during apnea or for assisted ventilation [4,11-13]. However, this method of preoxygenation is less effective in critically ill patients than NPPV (picture 1 and picture 2 and picture 3) or HFNO (figure 2) [30-32]. In clinical trials, HFNO and NPPV show similar effectiveness overall, with NPPV being slightly better in patients with severe hypoxemia [33]. Nevertheless, despite advanced preoxygenation with either HFNO or NPPV, severe oxygen desaturation occurs in a significant percentage (25 to 35 percent) of critically ill patients.

A systematic review and meta-analysis of HFNO for apneic oxygenation shows that apneic oxygenation has the greatest effect at reducing severe desaturation episodes in patients without significant hypoxemia as determined by the PaO₂/FiO₂ ratio [27]. In other words, the more hypoxemic a patient, the less effective apneic oxygenation is at maintaining a normal oxygen saturation.

Inadequate preoxygenation

●Management – In the absence of pulmonary disease, overcoming ambient air entrainment is the key step to preventing inadequate preoxygenation (denitrogenation). This can be accomplished by administering oxygen via any of the following measures:

•Flush-flow oxygen using a standard nonrebreathing mask – Open the valve on the wall regulator all the way until it stops to allow full flow. This is "flush-flow" rate oxygen.

•Tidal breathing 100 percent oxygen with a tight-fitting face mask – Use a tight-fitting nonrebreather mask, anesthesia circuit, or noninvasive ventilation. We do not recommend using a bag-valve mask for preoxygenating conscious patients, unless there are no other options, given the poor outcome data using this approach. When using noninvasive ventilation, set the FiO₂ to 1 (ie, 100 percent oxygen).

•HFNO – Use a commercially available high-flow system (eg, Optiflow cannula with heater and humidifier or AirVo system, Vapotherm). Use maximum flow for the selected system, with the FiO₂ at 1 (100 percent oxygen).

Keep the source of oxygen on until the patient is apneic (if using any mask system) and throughout the intubation if using HFNO to prevent renitrogenation from agonal breaths [25].

●Monitoring – End-tidal CO₂ is not specific for monitoring denitrogenation. However, monitoring end-tidal O₂ can be useful if available. The goal end-tidal O₂ is >80 percent, which represents complete denitrogenation. Measuring end-tidal O₂ can be difficult given the high inflow of oxygen, and the necessary equipment is rarely available outside the operating room.

●Apneic oxygenation – Apneic oxygenation is most likely to be helpful in keeping the oxygen saturation above 90 percent when denitrogenation is the rate-limiting step. We suggest the following approaches:

•If HFNO was used for preoxygenation, keep the HFNO in place during the apneic period at maximum flow and FiO₂. This is our preferred approach.

•While HFNO is a superior method, if it is not available, standard nasal cannula may be used. One reasonable approach is to use a flow rate of 15 liters per minute while the patient is awake, but to increase this flush rate once induction medication is administered.

Reduced functional residual capacity — FRC is reduced by any source of external compression (eg, obesity, ascites, pleural effusion, pneumothorax, hemothorax, term pregnancy) or by parenchymal volume loss (eg, pneumonia, ARDS, heart failure). The safe apnea period during RSI is shorter in proportion to the reduction in FRC.

●Management – In patients with reduced FRC, increasing end-expiratory lung volume is the priority during preoxygenation. This can be accomplished with the following measures:

•Reduce external compression

Upright positioning – Upright positioning can reduce external compression in patients with obesity, pregnancy, ascites, or comparable conditions. If spinal precautions are not necessary, we place the patient in a 30-degree head-up position [32,34,35].

If spinal precautions are necessary, modest upright positioning can be implemented by placing the bed in reverse Trendelenburg (bed flat but angled with the patient's head up (figure 3)).

Drain pleural space – Chest imaging, either by ultrasound or radiograph, may reveal a pneumothorax, hemothorax, or pleural effusion. Any clinically significant pneumothorax must be evacuated prior to intubation. When possible, a hemothorax or pleural effusion can be drained prior to intubation to improve preoxygenation and safe apnea time.

•Maximize alveolar recruitment – The following methods can be used to improve alveolar recruitment:

Noninvasive respiratory support – HFNO (if using a nasal high-flow system) and positive end-expiratory pressure (PEEP; via PEEP valve, continuous positive airway pressure, or noninvasive positive-pressure ventilation [NIPPV]) increase end-expiratory lung volume and should be used in critically ill patients when reduced FRC is suspected and no contraindications are present (table 3).

If using HFNO, use the maximum flow rate and keep the cannula in place during apnea.

If using PEEP, increase it to 10 to 12 cm H₂O during preoxygenation.

Invasive respiratory support – Induction followed by placement of a supraglottic airway device allows for higher airway pressures to be applied and helps to maintain a patent airway. This approach is appropriate for patients who are unconscious or breathing inadequately and meet the following criteria:

-Noninvasive respiratory support is contraindicated or not possible

-Upright positioning is not possible

-Mask ventilation with a PEEP valve is inadequate

When using this approach, waveform capnography is essential to ensure adequate ventilation, as patients are likely to desaturate quickly while the supraglottic airway is placed.

●Monitoring – End-tidal O₂ can be helpful for monitoring denitrogenation but is not helpful for measuring changes in FRC. If time is available, imaging such as plain chest radiography or pleural ultrasonography may be useful to assess volume loss after implementing either of the airway interventions listed immediately above.

●Apneic oxygenation – Apneic oxygenation can be helpful in patients with reduced FRC to maintain open alveoli, although it may not keep the oxygen saturation above 90 percent. Effective preoxygenation is most likely achieved using noninvasive respiratory support, so we suggest the following approaches:

•If HFNO was used for preoxygenation, maintain it during apnea at maximum flow and FiO₂. Maximum flow is needed to maintain end-expiratory lung volume during apnea.

•If NIPPV was used for preoxygenation, HFNO may be added for apneic oxygen at maximum flow and FiO₂.

Low ventilation-perfusion ratio (high shunt) — A refractory, low ventilation-to-perfusion ratio (so-called ventilation-perfusion or V/Q mismatch) is the most severe obstacle to preoxygenation. Often, improving the FRC (as discussed above) improves ventilation-perfusion mismatch. However, in patients with a mismatch that is refractory to such interventions, pulmonary blood flow is shunted past poorly ventilated alveoli. This occurs most often in severe, acute hypoxemic respiratory failure, such as multifocal pneumonia, and ARDS.

●Patients at risk — The risk of oxygen desaturation from increased shunt is greatest in patients with the following:

•Anatomic shunt – Patients with right-to-left intracardiac shunt (eg, atrial septal defect) or intrapulmonary shunt (eg, arteriovenous malformation, hepatopulmonary syndrome).

•Physiologic shunt – Patients with ARDS, multifocal pneumonia, or pulmonary edema and associated hypoxemia refractory to advanced preoxygenation strategies.

•High respiratory effort – Patients with ARDS and a high inspiratory effort on noninvasive respiratory support present a unique clinical challenge. In these patients, high inspiratory effort leads to high transpulmonary pressures that are inhomogeneously spread throughout the lung parenchyma and amplified in areas of injury. This causes air with a higher oxygen concentration to be pulled from areas of high compliance (apices) to areas of low compliance (bases), the so-called pendelluft effect, which masks the degree of V/Q mismatch. During induction, this effect is lost, which can lead to immediate oxygen desaturation.

●Management

•Maximize functional residual capacity – Improving functional residual capacity is the key intervention for reducing a low ventilation-perfusion ratio. This can be accomplished using the interventions described above. (See 'Reduced functional residual capacity' above.)

When hypoxemia persists despite improved FRC, inhaled pulmonary vasodilators can reduce ventilation-perfusion mismatch. We suggest adding inhaled nitric oxide at 5 to 10 ppm (parts per million) through the noninvasive respiratory circuit.

•Modified approaches to RSI – If RSI remains the preferred strategy, as with patients with depressed mental status, there are two options:

Induction with apneic oxygenation using a high-flow nasal system – This approach can reduce the incidence and depth of desaturation compared with mask ventilation after preoxygenation and induction [28].

Induction and placement of a supraglottic airway – This approach requires induction with immediate insertion of a supraglottic airway such as a laryngeal mask airway or the i-gel that can maintain a high seal pressure. Placing the supraglottic airway immediately after induction eliminates any apnea time and provides increased PEEP during ventilation, thereby recruiting more alveoli and reducing ventilation-perfusion mismatch.

Once preoxygenation is complete and intubation is ready to be performed, a few options are available:

-The supraglottic airway can be removed just prior to laryngoscopy and intubation performed in the usual fashion.

-The supraglottic airway can be left in place and used as a conduit for intubation using a flexible endoscope.

This approach requires additional equipment and expertise with flexible endoscopy. In addition, removal of the supraglottic airway after intubation can be challenging and there is some risk of accidental extubation and loss of the airway. There are techniques for intubating around the supraglottic airway, but such approaches are outside the scope of this review.

Awake intubation – If the patient is cooperative on noninvasive respiratory support an awake intubation may be preferable to RSI in the patient with refractory hypoxemia [36]. (See "Awake tracheal intubation".)

●Apneic oxygenation – Apneic oxygenation may be helpful in patients with significant V/Q mismatch to prevent cardiopulmonary arrest from severe desaturation but is unlikely to maintain an oxygen saturation above 90 percent. We suggest using HFNO at maximum flow and FiO₂ (ie, 100 percent O₂).

Hypotension and shock — There is no single best approach or universal intervention for reducing the risk of cardiovascular collapse during intubation. Given the complexities of managing critical illness and hemodynamic instability, we advocate a principle-driven approach based on the patient's risk for decompensation.

Risk assessment — While risk stratification scores are in development (eg, HYPES score), we perform a rapid risk assessment based on the blood pressure and shock index [18]. A systolic blood pressure of <90 mmHg or a mean arterial pressure <65 mmHg constitutes significant hypotension. The shock index is heart rate divided by systolic blood pressure (HR/SBP), with the normal range usually from 0.5 to 0.7.

In older patients with systemic hypertension, a "normal" systolic blood pressure measurement (eg, 110 mmHg) may be present in the setting of shock. Likewise, in a healthy young person, a "normal" blood pressure may be present despite significant shock, as such patients can produce intense vasoconstriction thereby maintaining blood pressure. Thus, it is important to look for other signs of hypoperfusion, such as altered mental status and diminished capillary refill. The assessment of shock in adult trauma patients is reviewed in detail separately (table 4 and algorithm 1 and table 5). (See "Approach to shock in the adult trauma patient".)

Based on the initial risk assessment, patients fall into one of three categories:

●Low risk – Normal blood pressure and normal shock index.

●Moderate risk – Normal blood pressure but elevated shock index (>0.7).

Induction and the transition to positive pressure ventilation can tip these patients into cardiovascular collapse. They generally have high sympathetic tone from the response to stress, including ample endogenous catecholamines, resulting in tachycardia but a relatively preserved blood pressure.

●High risk – Low blood pressure and elevated shock index (>0.7).

These patients are already in shock and should be resuscitated prior to intubation.

Patients deemed to be moderate or high risk require further assessment and resuscitation prior to induction and intubation. Findings from the history, physical examination, and focused cardiac ultrasound imaging can characterize underlying hemodynamic abnormalities and help to tailor resuscitation interventions (table 6).

Determining underlying causes of hypotension — A focused history and physical examination should be performed to identify the most likely cause or causes of hypotension and possible shock (table 5). (See "Evaluation of and initial approach to the adult patient with undifferentiated hypotension and shock" and "Definition, classification, etiology, and pathophysiology of shock in adults".)

Principles and approaches to peri-intubation resuscitation — Interventions to resuscitate patients with hemodynamic instability who require intubation vary depending on their level of risk and underlying pathophysiology. Interventions may include:

●Intravenous (IV) fluid administration for volume depletion

●Blood product administration for acute blood loss

●Vasopressor administration for vasoplegia

●Inotropic agents for poor left ventricular contractility

●Induction agent selection and modification based on degree of hypotension

●Right ventricular (RV)-focused resuscitation for RV failure

Common conditions that cause or exacerbate hypotension following RSI and interventions to help prevent or manage them are summarized in the following table (table 7).

General interventions to prevent hemodynamic collapse regardless of individual risk or underlying pathophysiology have proven unsuccessful. In two randomized trials, empiric initiation of a rapid 500 mL fluid bolus prior to induction of all critically ill patients, or of patients preoxygenated with noninvasive ventilation or a bag-valve mask, did not lead to significant reductions in the rate of cardiovascular collapse [26,37]. Likewise, imprecise administration of a vasopressor prior to intubation did not reduce rates of cardiovascular collapse, according to a secondary analysis of trial data [38].

Assess volume status and replete as needed — Replacing volume loss is often the first step in mitigating risk. However, as shown in randomized trials, initiating a 500 mL IV fluid bolus does not reduce risk in all patients [26,37]. Some patients are not volume-depleted or do not respond to volume. In other patients, 500 mL of isotonic crystalloid may be insufficient, particularly if there is significant intravascular volume loss. Therefore, fluid status should be assessed and the necessary fluid amounts provided as indicated. There are several ways to assess volume responsiveness, each with limitations, especially in the unintubated patient [39]:

●Passive leg raise – The passive leg raise test involves changing the patient's position from semi-upright to flat with their legs elevated to 45 degrees, thereby increasing venous return and preload. A quantitative increase in cardiac output of 5 percent as indicated by an increase in ETCO₂, or a qualitative increase as indicated by an improvement in blood pressure and heart rate, in response to this maneuver strongly predicts a positive response to fluid boluses. However, the test may be inaccurate in critically ill patients with complications (eg, hypovolemia, elevated intra-abdominal pressure). (See "Novel tools for hemodynamic monitoring in critically ill patients with shock", section on 'Passive leg raising or fluid bolus challenge'.)

●Arterial waveform analysis – In patients with an arterial catheter, analysis of waveform changes during respiration can help to predict responsiveness to fluid resuscitation. Changes of 10 percent or more in pulse pressure or stroke volume suggest volume depletion. Also, changes in systolic pressure with respiration, or narrowing of respiratory-related fluctuations when a passive leg raise is performed, indicate likely volume responsiveness. However, in critically ill patients, such analysis may be limited in the setting of dysrhythmia, RV dysfunction, alterations in arterial tone, and other pathology. (See "Novel tools for hemodynamic monitoring in critically ill patients with shock", section on 'Pulse contour analysis (fluid responsiveness)'.)

●Point-of-care ultrasound – Doppler ultrasound can be used to assess fluid responsiveness by determining stroke volume through the left ventricular outflow tract (LVOT). This approach is not affected by changes in arterial tone. It also provides insight into ventricular contractility and can quantify stroke volume and cardiac output.

The combination of changes in stroke volume with respiration (fluctuations in the LVOT velocity time integral [VTI]), and measurements of blood pressure/pulse pressure can help to guide interventions, particularly in moderate-risk patients (table 6):

•Stroke volume (LVOT VTI) variation high, pulse pressure low – This phenotype is commonly seen in states of high systemic vascular resistance and volume depletion (eg, hemorrhagic shock). In these patients, resuscitation with an appropriate fluid (eg, blood products) and re-evaluation should be the next step.

•Stroke volume (LVOT VTI) variation low, pulse pressure high – This phenotype (especially if the diastolic blood pressure is low and VTI is high) is commonly seen in patients with vasoplegia (eg, gram-negative sepsis). In these patients, a vasopressor is needed to increase arterial tone prior to or in combination with fluid resuscitation.

Mixed phenotypes, such as large stroke volume variation with a wide pulse pressure, are common. It is also common to have more than one problem, such as low overall stroke volume because of poor contractility. Such patients are typically high risk (hypotensive and high shock index). The principles outlined below inform the management of such patients.

Reduce vasoplegia to increase intravascular (stressed) volume — In vasodilatory states (eg, gram-negative septic shock), the tone of the venous and arterial systems is low and fluid resuscitation alone is unlikely to be effective. A vasopressor is necessary to reduce vasoplegia and restore cardiac output.

In such cases, we advocate using continuous vasopressor infusions rather than bolus dosing. Norepinephrine is an excellent first choice. A number of factors determine appropriate vasopressor selection and dosing, which is discussed in detail separately. (See "Use of vasopressors and inotropes".)

IV fluid that does not increase intravascular pressure is "unstressed." Normally, unstressed fluid is maintained primarily in the high-capacitance venous system. When a higher cardiac output is needed, venoconstriction converts that fluid from unstressed to stressed, thereby increasing venous return. In critically ill patients with vasoplegia, vasopressors may be necessary to improve vascular tone and may also increase responsiveness to fluids.

Augment left ventricular performance — If the left ventricle is contracting or relaxing ineffectively, the patient is at risk of cardiovascular collapse during RSI.

●Poor ejection fraction/contractility – Poor contractility of the left ventricle reduces the ejection fraction and stroke volume. Such a compromised left ventricle is unable to increase stroke volume in response to any drop in afterload caused by induction and intubation. In these patients, we recommend starting a continuous norepinephrine infusion.

If contractility does not improve with norepinephrine, a dedicated inotrope (eg, dobutamine) should be administered in addition to norepinephrine. In patients with severe acidemia or bradycardia, epinephrine may be a good second agent.

●Poor relaxation/restrictive physiology – In this setting, stroke volume is already maximized in response to a high afterload. Any drop in preload or afterload with induction and intubation risks cardiovascular collapse. We recommend a norepinephrine infusion to maintain afterload, followed by fluid loading if clinically appropriate to maintain preload. Fluid loading is performed by giving isotonic fluid in 250 mL aliquots. Additional boluses are given based on changes in blood pressure and cardiac output.

Protect the right ventricle — When RV dysfunction is identified or suspected, appropriate measures to mitigate risk should be taken. We suggest the approach outlined below.

Under normal conditions, the RV facilitates cardiac output by increasing venous return to the left side of the heart. A dysfunctional or failed RV compromises cardiac output, and this tendency is exacerbated by any drop in preload caused by induction or by any increase in afterload caused by intubation. Studies of intubation in critically ill patients report a significantly increased risk of hemodynamic instability or cardiac arrest in patients with moderate or severe RV dysfunction [18,21].

●Evaluate RV function – Examination findings suggesting RV dysfunction include:

•Jugular venous distension

•Pedal edema

•Hepatic congestion and ascites

Clinicians should be very concerned about RV failure in patients with any of these findings, particularly those who are already hypotensive or at moderate risk for peri-intubation decompensation based on the initial assessment. (See 'Risk assessment' above.)

When possible, focused cardiac ultrasound is most helpful in determining the presence and severity of RV dysfunction and potential reserve. Important findings include:

•Parasternal short axis view – A dilated RV with septal flattening is a concerning finding (image 1 and movie 1). Septal flattening during systole only suggests pressure overload, but septal flattening throughout the cardiac cycle suggests pressure and volume overload.

•Apical four chamber view – The tricuspid annular plane systolic excursion (TAPSE), using M-mode through the lateral tricuspid annulus, measures how far the tricuspid annulus moves during RV contraction. The more strain the RV is under, the less it moves during systole and the TAPSE measurement falls (image 2 and movie 2). This is evidence of RV systolic strain.

A TAPSE <16 mm is abnormal. Tissue Doppler through the same location reveals the speed of the contraction, essentially the contractile force, reflected in the isovolumetric contraction velocity (IVV) before the pulmonic valve opens, or the S' after the pulmonic valve opens. Patients with these findings are often tachycardic in an attempt to maintain cardiac output, and the IVV and S' waves are blended. If either wave is <10 cm/second, that is indicative of poor contractile force.

●Management based on cardiac ultrasound findings – We recommend the following approach to resuscitation in patients with RV dysfunction and the following findings on echocardiogram (table 6):

•Right atrium and inferior vena cava dilated WITHOUT septal flattening on diastole: Administer an infusion of norepinephrine or vasopressin (phenylephrine should not be used).

•Right atrium and inferior vena cava dilated with septal flattening on diastole: Initiate diuresis with a medication such as furosemide (eg, 40 mg IV). Start an infusion of norepinephrine or vasopressin. Re-evaluate septal movement if time permits.

•TAPSE <16 mm + IVV/S' <10 cm: Administer an inhaled pulmonary vasodilator. We use inhaled nitric oxide at 40 to 80 ppm that is bled inline into the circuit used for preoxygenation. Nitric oxide can be added to the circuit of any oxygen delivery source, including nasal cannula, nonrebreather face mask, NIPPV, or invasive mechanical ventilation. If the dose of inhaled nitric oxide leads to hypoxemia, it can be reduced to 20 ppm.

•TAPSE/right ventricular systolic pressure <0.31: These patients are in cardiogenic shock and may need mechanical circulatory support, such as an Impella device or extracorporeal membrane oxygenation, if the hemodynamics cannot be improved with the resuscitation strategies described above.

Mitigate secondary effects of induction — Propofol, etomidate, and ketamine are the primary induction agents used for RSI and are of particular interest when considering the risk for cardiovascular collapse. No induction agent is safe for all patients. Relying on the secondary or indirect effects of any induction agent to mitigate the risk of cardiovascular collapse caused by underlying physiologic abnormalities is unlikely to be successful.

Key principles for reducing the risks associated with induction agents include the following:

●Resuscitate adequately prior to intubation so secondary effects of the induction agent become less significant.

●Resuscitate in anticipation of the changes in hemodynamics that occur with induction and intubation, not simply to achieve adequate blood pressure prior to intubation.

●Reduce the dose of the induction agent in proportion to the degree of hemodynamic dysfunction.

Dose reduction may be important, particularly when giving ketamine or propofol. Giving a full dose can undermine previous resuscitation interventions. The reduction required will vary with the clinical situation, but a half dose or even one-third dose may be necessary. In such circumstances, one author often modifies RSI by giving small bolus doses of ketamine (eg, 20 mg) or propofol (eg, 10 mg) until the patient is unconscious, and then give a bolus of the full dose of the neuromuscular blocking agent.

Observational studies performed in critically ill patients report disparate outcomes with each of the induction agents. As an example, following intubation, ketamine has been associated with an increased risk of hypotension [40,41], a decreased risk of hypotension [42], and no difference compared with etomidate [43,44]. Several randomized trials comparing ketamine and etomidate have found no clinically significant difference [45-47]. One randomized trial reported higher seven-day mortality with etomidate but no difference at 28 days, and secondary hemodynamic outcomes all trended worse for ketamine [48]. Neither mixing ketamine and propofol ("ketofol") [49] nor reducing the dose of etomidate is associated with an overall reduction in postintubation hypotension [50].

Combination of hypoxemia and hypotension — Simultaneous hemodynamic instability and compromised gas exchange can coexist (eg, pneumonia complicated by septic shock). Such patients are at high risk for decompensation during intubation and are often difficult to resuscitate. Often, these two problems create a feedback loop, each exacerbating the severity of the other. As an example, hypoxemia can worsen hemodynamics, which increases dead space, further impairing gas exchange.

We recommend the following stepwise approach but recognize these interventions are often performed in parallel:

●Address hypoxemia first – Optimize preoxygenation. (See 'Hypoxemia' above.)

●Characterize and remediate the underlying hemodynamics. (See 'Hypotension and shock' above.)

Severe metabolic acidosis — In addition to treatment of the underlying cause of metabolic acidosis, the following interventions may reduce the risk of serious adverse events during intubation:

●Preoxygenation with high flow rates (eg, HFNO or NIPPV)

●Bicarbonate administration (no proof of benefit during peri-intubation period)

●Awake intubation (when indicated)

●Mechanical ventilation that matches the patient's minute ventilation requirement

Severe metabolic acidosis increases the work of breathing and leads to high minute ventilation. This makes preoxygenation challenging with a standard facemask due to dilution with ambient air. Therefore, it is important to perform preoxygenation with higher flow rates that come closer to meeting the inspiratory demands of these patients. This can only be achieved with noninvasive respiratory support, either HFNO or NIPPV. (See "Noninvasive ventilation in adults with acute respiratory failure: Benefits and contraindications" and "Noninvasive ventilation in adults with acute respiratory failure: Practical aspects of initiation".)

Bicarbonate administration is controversial and there are no clear data to demonstrate benefit, even short-term, or to prove safety. During RSI, the PaCO₂ can rise rapidly during the first minute of apnea, generally 10 to 13 mmHg, dropping the pH between 0.1 and 0.2 [51,52]. Such a drop in pH in a patient with severe metabolic acidosis may be sufficient to precipitate cardiac arrest. Treating the underlying cause of metabolic acidosis is paramount for reducing this risk.

In addition, despite the absence of supporting scientific evidence, some clinicians empirically administer bicarbonate. They believe bicarbonate may help to stabilize a pH that is consistently below seven despite resuscitation and administer it prior to intubation with the intention of achieving a higher pH during intubation, but at the cost of a higher acid burden subsequently. The use of bicarbonate as an adjunct therapy for severe metabolic acidosis is reviewed in detail separately. (See "Approach to the adult with metabolic acidosis" and "Bicarbonate therapy in lactic acidosis".)

Awake intubation is another possible strategy for patients who cannot tolerate any apnea or have minute ventilation requirements that are difficult to meet with mechanical ventilation, provided that oxygenation can be maintained during the procedure. (See "Awake tracheal intubation".)

In either case, once intubation is achieved, mechanical ventilation using appropriate settings is critical. Clinicians must attempt to match the patient's minute ventilation requirement, especially if a long-acting paralytic was used for intubation. A spontaneous breathing mode such as pressure support ventilation may be best once the patient is able to breathe spontaneously. (See "Mechanical ventilation of adults in the emergency department".)

If a spontaneous breathing mode is not possible, patients generally require mechanical ventilation with high tidal volumes and respiratory rates. Clinicians must watch closely for breath stacking and air trapping. In this setting, the principles of lung protective ventilation typically cannot be followed, so it becomes critical to treat the underlying acidosis aggressively to allow the introduction of lung protective strategies as soon as reasonably possible.

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Airway management in adults".)

SUMMARY AND RECOMMENDATIONS

●Definitions and epidemiology – The anatomically difficult airway is one in which anatomic characteristics (eg, recessed mandible, swollen tongue) make it difficult to perform tracheal intubation, mask ventilation, or rescue oxygenation. The physiologically difficult airway is one in which underlying physiologic abnormalities (eg, septic shock, heart failure) increase the risk of rapid oxygen desaturation, cardiovascular collapse, or both during the peri-intubation period. Intubation of critically ill patients is associated with serious adverse events in approximately 20 to 30 percent of cases. (See 'Terminology and epidemiology' above.)

Factors associated with increased risk of peri-intubation decompensation are reviewed in the text and include:

•Preintubation hypotension or hypoxemia

•Older age

•Sepsis

•Cardiac dysfunction

●Causes and prevention of peri-intubation hypoxia – These causes include inadequate preoxygenation, reduced functional residual capacity (FRC), and low ventilation-perfusion ratio (large shunt):

•Inadequate preoxygenation (denitrogenation) occurs in patients with a high respiratory effort that dilutes supplemental oxygen with ambient air (eg, multifocal pneumonia, sepsis). Interventions include delivering high-concentration oxygen (FiO₂ of 1) by high flow, while ensuring a tight mask seal. Apneic oxygenation is an important related technique for reducing the risk of oxygen desaturation. (See 'Preoxygenation (denitrogenation)' above and 'Inadequate preoxygenation' above.)

•Diminished FRC is caused by external compression (eg, obesity, pleural effusion, pneumothorax) or by functional parenchymal volume loss (eg, pneumonia, acute respiratory distress syndrome, heart failure). FRC provides the oxygen reservoir during the apneic phase of rapid sequence intubation (RSI). Interventions include upright positioning, draining pleural spaces, and maximizing alveolar recruitment with noninvasive (eg, high-flow nasal oxygen [HFNO]) or invasive respiratory support. (See 'Reduced functional residual capacity' above.)

•Refractory, low ventilation-to-perfusion ratio (aka V/Q mismatch) is the most severe obstacle to preoxygenation. In addition to interventions to increase FRC, management may include inhaled pulmonary vasodilators (eg, nitric oxide) and modified approaches to RSI, including induction with apneic oxygenation using HFNO or induction and a supraglottic airway, or awake intubation. (See 'Low ventilation-perfusion ratio (high shunt)' above.)

●Causes and prevention of peri-intubation hypotension – Clinicians should identify underlying causes and assess the risk associated with hypotension. Patients with elevated shock index and "normal" or low blood pressure are at elevated risk of peri-intubation cardiovascular collapse. (See 'Hypotension and shock' above.)

Interventions for patients with hemodynamic instability who require intubation vary depending on their level of risk and underlying pathophysiology and may include:

•Isotonic intravenous fluid for volume depletion

•Blood products for acute blood loss

•Vasopressor (eg, norepinephrine) infusion for vasoplegia

•Inotropic agents for poor left ventricular contractility

•Induction agent modification (reduced dose)

●Right ventricle-focused resuscitation (when required)

Awake intubation is another approach should the clinician determine that RSI poses too great a risk. (See "Awake tracheal intubation".)

●Point-of-care ultrasound (POCUS) assessment – POCUS is an important tool for determining the type and severity of hemodynamic dysfunction (table 6). Details of the POCUS assessments and related interventions to mitigate problems are found in the text:

•Fluid status (see 'Assess volume status and replete as needed' above)

•Vasoplegia (see 'Reduce vasoplegia to increase intravascular (stressed) volume' above)

•Right atrial filling & right ventricle function (see 'Protect the right ventricle' above)

•Left ventricle function (see 'Augment left ventricular performance' above)