Overview of initiating invasive mechanical ventilation in adults in the intensive care unit

INTRODUCTION — There are several indications for the initiation of invasive mechanical ventilation in the intensive care unit (ICU) (table 1). The common modes of ventilation, initial settings, and supportive care for intubated patients are discussed in this topic review. Intubation and complications of invasive mechanical ventilation are described separately. (See "Ventilator-induced lung injury" and "Clinical and physiologic complications of mechanical ventilation: Overview" and "Direct laryngoscopy and endotracheal intubation in adults" and "Rapid sequence intubation in adults for emergency medicine and critical care" and "The decision to intubate" and "Complications of the endotracheal tube following initial placement: Prevention and management in adult intensive care unit patients".)

DEFINITION — Invasive mechanical ventilation is defined as the delivery of positive pressure to the lungs via an endotracheal or tracheostomy tube.

During mechanical ventilation, a predetermined mixture of air (ie, oxygen and other gases) is forced into the central airways and then flows into the alveoli. As the lungs inflate, the intra-alveolar pressure increases. A termination signal (usually flow or pressure) eventually causes the ventilator to stop forcing air into the central airways and the central airway pressure decreases. Expiration follows passively, with air flowing from the higher pressure alveoli to the lower pressure central airways.

Noninvasive ventilation (NIV) is delivered through an alternative interface, usually a face mask. Patient selection and indications for NIV are discussed in detail separately. (See "Noninvasive ventilation in adults with acute respiratory failure: Benefits and contraindications", section on 'Patients likely to benefit'.)

INDICATIONS — Invasive mechanical ventilation is most often used to fully or partially replace the functions of spontaneous breathing by performing the work of breathing and gas exchange in patients with respiratory failure (table 1 and table 2 and table 3) [1]. Invasive mechanical ventilation may also be useful in those who require airway protection to reduce the risk of aspiration (eg, depressed mental status from an overdose, patients with variceal bleeding). Importantly, regardless of the indication, invasive mechanical ventilation should be considered early in the course of illness and should not be delayed until the need becomes emergent. (See "Mechanical ventilation of adults in the emergency department" and "Measures of oxygenation and mechanisms of hypoxemia" and "The evaluation, diagnosis, and treatment of the adult patient with acute hypercapnic respiratory failure".)

SELECTING AN INITIAL MODE — Commonly selected initial modes and factors that influence mode selection are discussed in this section. Further details on the physiology that underlies each mode are provided separately. (See "Modes of mechanical ventilation".)

Commonly used modes — There is no universal initial mode of invasive mechanical ventilation (table 4) that is ideal for all patients. However, common initial modes which are suitable for most patients include:

●Volume-limited assist control ventilation

●Pressure-limited assist control ventilation

●Synchronized intermittent mandatory ventilation (SIMV) with pressure support ventilation (SIMV-PSV)

Pressure support ventilation (PSV) alone is uncommonly used as an initial mode of ventilation but commonly used during weaning (see "Initial weaning strategy in mechanically ventilated adults", section on 'Choosing a weaning method'). Continuous (formerly known as "controlled") mechanical ventilation (CMV; volume-limited CMV or pressure-limited CMV), intermittent mandatory ventilation (IMV), and airway pressure release ventilation (APRV) are not typically used as initial modes of ventilation. Adaptive support ventilation (ASV), and neurally adjusted ventilatory assist ventilation (NAVA) are investigational modes and high frequency mechanical ventilation is not recommended in adults. (See "Modes of mechanical ventilation".)

Regardless of the initial mode selected, the mode of ventilation can often be changed when a patient demonstrates intolerance of selected mode or demonstrates the signs of dyssynchrony. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Treat dyssynchrony'.)

The modes of mechanical ventilation are distinguished from each other by the types of breaths that they deliver (table 4). In brief, the delivery of breaths is typically either volume-limited or pressure-limited:

●Volume-limited – Volume-limited breaths can be ventilator-initiated (also known as volume-controlled or volume-cycled [VC]) or patient initiated (also known as volume-assist [VA]). VC or VA breaths deliver a predetermined tidal volume at a set ventilator rate such that a minimum minute ventilation (tidal volume x respiratory rate) is guaranteed. Each tidal volume is delivered at a set inspiratory flow rate and inspiration is terminated once the set tidal volume has been delivered. In this mode, airway pressure is dependent on the airway resistance, lung compliance, and chest wall compliance. Modes of mechanical ventilation used in the ICU that can deliver VC or VA breaths include volume-limited assist control and volume-limited SIMV. Volume-limited continuous mechanical ventilation (VC-CMV) is not generally used in the ICU. Pressure-regulated volume control ventilation (PRVC) is being increasingly used. In PRVC, a set tidal volume is targeted by varying airway pressure resulting in variable inspiratory flow. Further details are provided separately. (See "Modes of mechanical ventilation", section on 'Volume-limited ventilation'.)

●Pressure limited – Pressure-limited breaths can be ventilator-initiated (also known as pressure-control or pressure-cycled [PC]) or patient initiated (also known as pressure-assist [PA]). In PC or PA breaths, the flow of air into the lung is determined by a set pressure limit and the rate is determined by a set ventilator rate. Inspiration is terminated once the set inspiratory time has elapsed. The tidal volume for each breath is variable and related to compliance, airway resistance, and tubing resistance. A consequence of the variable tidal volume is that a specific minute ventilation cannot be guaranteed. Modes of mechanical ventilation used in the ICU that deliver PC or PA breaths include pressure-limited assist control and pressure-limited SIMV. Pressure-limited continuous mechanical ventilation (PC-CMV) is not generally used in the ICU. Further details are provided separately. (See "Modes of mechanical ventilation", section on 'Pressure-limited ventilation'.)

●Pressure support – Spontaneous breathing can be supported to a set pressure limit; such breaths are called "pressure support (PS) breaths." The ventilator provides the driving pressure for each spontaneous breath, which determines the maximal airflow rate. Inspiration is terminated once the inspiratory flow has decreased to a predetermined percentage of its maximal value. (See "Modes of mechanical ventilation", section on 'Pressure support'.)

●Other modes that are not generally used as the initial mode but may be used during the course of mechanical ventilation are described separately. (See "Modes of mechanical ventilation", section on 'Adaptive support ventilation' and "Modes of mechanical ventilation", section on 'Neurally adjusted ventilatory assist ventilation' and "Modes of mechanical ventilation", section on 'High frequency ventilation'.)

Factors influencing the choice — Factors that influence the initial mode chosen include:

●Level of support needed – The level of ventilator support is the proportion of the patient's ventilatory needs that are met by the ventilator. The level of support is determined by the following:

•The mode – Generally speaking, among the modes used in the ICU, volume- or pressure-limited modes that use assist control functions tend to provide the most support (ie, resting respiratory muscles while simultaneously minimizing atrophy). In contrast, pressure support tends to provide the least support and is associated with a greater work of breathing.

•The indication for mechanical ventilation – In general, patients who are mechanically ventilated for respiratory failure need more support than those ventilated for airway protection while those with severe respiratory failure need more support than those with mild respiratory failure.

•The selected settings – Selected settings can be adjusted to provide an increased proportion of reduced support. (See 'Settings' below.)

●Reason for mechanical ventilation – The indication for mechanical ventilation may influence the mode. For example, patients with acute respiratory distress syndrome (ARDS) are typically placed on low tidal volume ventilation (LTVV) delivered using volume-limited assist control ventilation or pressure-limited assist control ventilation. In contrast, a patient who is mechanically ventilated short-term for airway protection may be suitable for several additional modes including SIMV-PSV as well as PSV alone. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Volume- versus pressure-limited mode'.)

●Presence of airflow limitation – In patients with active airflow limitation such as severe chronic obstructive lung disease [COPD], acute asthma, or acute COPD exacerbation, volume-limited modes of ventilation are commonly used (eg, volume-limited assist control ventilation, SIMV-PSV). In contrast, pressure support or pressure-limited modes including APRV are generally avoided. (See "Invasive mechanical ventilation in acute respiratory failure complicating chronic obstructive pulmonary disease" and "Invasive mechanical ventilation in adults with acute exacerbations of asthma".)

●Presence of an air leak – In patients with a prolonged, significant air leak (eg, from pneumothorax or lung surgery) pressure-limited ventilation or SIMV-PSV, or even PSV alone, are preferred in order to limit additional barotrauma and worsening of the air leak.

●Concern for elevated intracranial pressure (ICP) – While volume-limited modes are frequently used in patients with an elevated ICP (eg, traumatic brain injury or stroke), pressure-limited modes are popular due to the theoretical concern that elevated intrathoracic pressures hamper venous return from the brain and therefore worsen ICP. However, there are no data in this population to suggest that one mode is superior to the other and practice varies widely. (See "Management of acute moderate and severe traumatic brain injury", section on 'Ventilation'.)

●Paralysis – For patients who are paralyzed or heavily sedated (ie, those in whom the initiation of a spontaneous breath is limited), PSV is contraindicated and assist controlled modes (volume- or pressure-limited) are typically used, provided the set minute ventilation is adequate.

●Others – Clinician familiarity and institutional preferences often dictate common modes used.

SETTINGS — Once the mode is selected, there are numerous parameters that need to be set, many of which depend upon the mode selected.

Mode-specific settings — Included in this section are suggested initial settings for the commonly used modes of ventilation (table 5). However, it should be noted that settings should be tailored to the patient who is being ventilated. Importantly, it is critical that the patient be monitored and settings adjusted according to clinical and, when available, blood gas findings. (See 'Follow-up' below.)

Volume-limited assist control ventilation — For patients in whom volume-limited assist control ventilation is chosen as the initial mode of ventilation, we typically use the following initial settings:

●Tidal volume – Tidal volume is typically set at 6 mL per kg predicted body weight (PBW); 4 to 8 mL/kg PBW for patients with acute respiratory distress syndrome (ARDS) and 6 to 8 mL/kg PBW for patients who do not have ARDS. (See 'Tidal volume' below.)

●Ventilator rate – Ventilator rate is typically set at 12 to 16 breaths per minute; higher rates may be necessary for patients with ARDS (eg, ≤35 breaths per minute). (See 'Ventilator rate' below.)

●Positive end-expiratory pressure (PEEP) – PEEP is typically set at 5 cm H₂O with subsequent adjustments made according to the fraction of inspired oxygen (FiO₂) (table 6). (See 'Positive end-expiratory pressure' below.)

●FiO₂ – FiO₂ is set to maintain the peripheral oxygen saturation (SpO₂) between 90 and 96 percent. (See 'Fraction of inspired oxygen' below.)

●Inspiratory flow – Inspiratory flow is typically set at 40 to 60 L/minute with a ramp pattern to target an inspiratory:expiratory (I:E) ratio of approximately 1:2 to 1:3. Higher rates up to 75 L/minute that decrease the I:E ratio are appropriate in patients with airflow obstruction. (See 'Flow rate and pattern' below.)

●Trigger sensitivity – Typical values are 2 L/minute when flow-triggering is used or -1 to -2 cm H₂O when pressure-triggering is used. Pressure-triggering should not be used when auto-PEEP is suspected. (See 'Trigger sensitivity' below.)

A modified version of volume-limited ventilation known as pressure-regulated volume control (PRVC) is increasingly used in some ICUs. PRVC is a form of volume control ventilation where tidal volume is set and the applied airway pressure changes to attain the target tidal volume (resulting in variable inspiratory flow). In PRVC, clinicians set the inspiratory time instead of flow in addition to the desired tidal volume and the other parameters listed above. The inspiratory time should target an I:E ratio of 1:2 to 1:3.

Pressure-limited assist control ventilation — For patients in whom pressure-limited assist controlled ventilation is chosen as the initial mode of ventilation, we typically set the inspiratory pressure level to target an approximate tidal volume (eg, 4 to 8 mL/kg PBW for a patient with ARDS) and the inspiratory time is set to deliver an I:E ratio of 1:2 to 1:3 (typically one second). The FiO₂, ventilator rate, applied PEEP, and trigger sensitivity are similar to those of volume-limited-assist control ventilation (VC-AC). (See 'Volume-limited assist control ventilation' above.)

The initial inspiratory pressure set varies depending upon lung compliance, airway resistance, and tubing resistance, but in general, acceptable target tidal volumes may be reached with inspiratory pressure levels between 12 and 25 cm H₂O. However, the clinician should bear in mind that the addition of the applied PEEP to a set inspiratory pressure increases peak airway pressure and may further increase the risk of barotrauma (eg, patient with a set inspiratory pressure level of 20 cm H₂O and an applied PEEP of 10 cm H₂O will have a peak airway pressure of 30 cm H₂O).

Synchronized intermittent mandatory ventilation with pressure support ventilation (SIMV-PSV) — For patients in whom SIMV-PSV is chosen as the initial mode of ventilation, we typically use similar settings to VC-AC with the addition of pressure support (eg, 5 to 10 cm H₂O) for spontaneous breaths taken by the patient above the set rate. The pressure support can be subsequently increased as needed for patient comfort and reduced when mitigation of respiratory alkalosis is required.

Others — Should PSV alone be chosen, we generally increase the pressure support level (typically 5 to 15 cm H₂O but maybe up to 20 cm H₂O) until the patient's respiratory rate falls below 30 breaths per minute (a ventilator rate is not set for PSV) and a targeted tidal volume is achieved (eg, 4 to 8 mL/kg PBW). The FiO₂, applied PEEP, inspiratory flow rate, and trigger sensitivity are similar to those of VC-AC. (See 'Volume-limited assist control ventilation' above.)

General

Tidal volume — The tidal volume is the amount of air delivered with each breath. How volume is delivered differs with the mode of ventilation:

●During volume-limited ventilation, the tidal volume is set by the clinician and remains constant. Newer ventilators are now also capable of a PRVC (also known as VCPlus) mode, which provides variable inspiratory pressures to target a specific tidal volume. Although the clinician sets a target tidal volume, the actual volumes delivered to the patient may vary breath to breath based on the algorithm being used. (See 'Commonly used modes' above and "Modes of mechanical ventilation", section on 'Volume-limited ventilation'.)

●During pressure-limited ventilation, the clinician chooses pressure settings to achieve an "approximate" tidal volume. However, the amount of tidal volume generated by the inspiratory pressure is also determined by the lung compliance and resistance of the ventilator tubing. Thus the tidal volume is variable. (See 'Commonly used modes' above and "Modes of mechanical ventilation", section on 'Pressure-limited ventilation'.)

There is no universal tidal volume that is ideal for every patient. In our practice, we avoid volumes ≥10 mL/kg of PBW (ie, ideal body weight (calculator 1 and calculator 2)) and typically start with tidal volumes of 6 mL/kg PBW (range 4 to 8 mL/kg PBW depending upon the indication for mechanical ventilation). Tidal volumes higher than 10 mL/kg were common in the past but it is now known that large tidal volumes increase morbidity (eg, barotrauma or ventilator-associated lung injury [2,3]) and that low tidal volume ventilation (LTVV) has a mortality benefit in patients with acute respiratory distress syndrome (ARDS); the benefits in non-ARDS patients is less clear.

●Acute respiratory distress syndrome (ARDS) – In ARDS, we typically set the initial tidal volume at 6 mL/kg PBW (range 4 to 8 mL per kg PBW) since randomized trials have shown improved mortality with LTVV in this population (table 6 and table 7 and table 8) [4]. Data to support this strategy are provided separately. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Low tidal volume ventilation: Initial settings'.)

●Non-ARDS patients – The optimal tidal volume for patients who are mechanically ventilated for reasons other than ARDS is unclear. Based upon limited data, a tidal volume between 6 to 8 mL per kg of PBW is reasonable (table 8 and table 7). Occasionally, volumes between 8 and 10 mL/kg PBW are appropriate (eg, patients with double triggering).

•Medical – In this group of patients (eg, patients intubated for airway protection, patients who are not immediately postoperative, patients admitted to medical ICUs for nonsurgical reasons), we typically use an initial tidal volume of approximately 6 to 8 mL per kg of PBW and avoid large volumes (eg, 10 mL per kg of PBW).

One meta-analysis of 15 randomized trials and five observational studies reported that patients ventilated with a tidal volume approximately 6 mL/kg had a lower risk of lung injury (RR 0.33, 95% CI 0.23-0.47) and mortality (RR 0.64, 95% CI 0.46-0.89) compared with patients ventilated with higher tidal volumes [5]. A second meta-analysis of seven studies found that fewer patients on low tidal volume ventilation (≤7 mL/kg) progressed to ARDS, developed pneumonia, or died than patients receiving higher tidal volumes (>7 mL/kg) [6]. However, both analyses were limited by the methodologic limitations of some of the randomized trials, the inclusion of several observational studies and the wide heterogeneity of the included populations (eg, the inclusion of intraoperative patients).

Another randomized trial (PReVENT) compared low tidal volume (<6 mL/kg PBW; target of 4 mL/kg PBW) with intermediate tidal volume (10 mL/kg PBW) ventilation strategies in patients with respiratory failure other than ARDS who were predicted to require ventilation for longer than 24 hours [7]. No difference between the groups in ventilator-free days, mortality, length of stay, or rate of infections, ARDS, or pneumothorax was reported. However, by day 1, almost two-thirds of patients in the low tidal volume group received volumes >6 mL/kg PBW, 14 percent had a tidal volume >9.5 mL/kg PBW, and less than 25 percent achieved the target volume of 4 mL/kg PBW. Similarly, only one-quarter of patients in the intermediate tidal volume group received tidal volume ≥10 mL/kg PBW. Thus, insufficient differences between the achieved tidal volumes received by patients in both groups may have contributed to the lack of benefit.

•Surgical (ie, intraoperative, postoperative) – There are a paucity of data on the optimal intraoperative tidal volume for patients who are mechanically ventilated during surgery. However, many studies have shown a lower rate of postoperative pulmonary complications with the LTVV approach, the details of which are discussed separately. (See "Mechanical ventilation during anesthesia in adults", section on 'Lung protective ventilation during anesthesia'.)

Few studies have specifically examined the effect of tidal volume in the postoperative population and although 17 percent of patients in PReVENT were surgical ICU patients, no subgroup analysis was performed in that group [7].

Once set, the tidal volume can be subsequently increased or decreased incrementally to achieve the desired pH and arterial carbon dioxide tension (PaCO₂), and patients should be monitored for intrinsic PEEP (intrinsic- or auto-PEEP) and high airway pressure. Excessively high tidal volumes place the patient at risk of barotrauma, auto-PEEP, hyperventilation, and respiratory alkalosis while excessively low tidal volumes can result in atelectasis, hypoventilation, and respiratory acidosis. (See 'Follow-up' below.)

Ventilator rate — With the exception of pressure support ventilation, most forms of volume-limited and pressure-limited ventilation require a set ventilator rate. An optimal method for setting the ventilator rate has not been established. However, for most patients, an initial set ventilatory rate between 16 and 30 breaths per minute is reasonable. Importantly, the clinician should understand that for patients who breathe above the set rate, the true minute ventilation (respiratory rate x tidal volume) will be determined by the native respiratory rate and the actual delivered tidal volume; careful comparison of the set minute ventilation and the actual minute ventilation is important when interpreting gas exchange parameters. (See 'Follow-up' below.)

Several factors may additionally be taken into consideration when setting the rate:

●Mode – For patients receiving assist control modes, the ventilatory rate is typically set approximately four breaths per minute below the patient's native rate (eg, mid 20s for a patient breathing at 30 breaths per minute). For patients receiving synchronized intermittent mandatory ventilation, some experts set the rate to ensure that at least 80 percent of the patient's total minute ventilation is delivered by the ventilator. (See "Modes of mechanical ventilation", section on 'SIMV'.)

●ARDS – For patients with ARDS, higher rates may be needed in order to facilitate low tidal volume ventilation necessary for this population (eg, up to 35 breaths per minute) but most often rates are set between 14 and 22 breaths per minute. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Low tidal volume ventilation: Initial settings'.)

Over the course of ventilation, the ventilator rate can be incrementally increased or decreased to achieve the desired minute ventilation, pH and PaCO₂. Excessively high rates may be complicated by auto-PEEP (see "Positive end-expiratory pressure (PEEP)", section on 'Auto (intrinsic) PEEP') and respiratory alkalosis while excessively low rates may be complicated by respiratory acidosis; patients may continue to have respiratory acidosis despite optimization of their tidal volume and ventilator rate settings. In this situation, permissive hypercapnic ventilation is appropriate. (See "Permissive hypercapnia during mechanical ventilation in adults".)

Positive end-expiratory pressure — Extrinsic PEEP (ie, applied PEEP) is generally added to mitigate end-expiratory alveolar collapse. A typical initial applied PEEP is 5 cm H₂O. However, up to 24 cm H₂O may be used in patients undergoing low tidal volume ventilation for ARDS. A suggested adjustment strategy in patients with ARDS is shown in the table (table 6). (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Low tidal volume ventilation: Initial settings'.)

Optimal levels in non-ARDS patients are unknown but, in general, excessively high levels are avoided since elevated levels of applied PEEP can have adverse consequences, such as reduced preload (decreases cardiac output), elevated plateau airway pressure (increases risk of barotrauma), and impaired cerebral venous outflow (increases intracranial pressure). Excessively low levels should also probably be avoided to avert adverse effects, such as hypoxemia and atelectasis. Consistent with the practice of others, we typically choose PEEP of approximately 5 cm H₂O. Data that support this practice are provided separately. (See "Positive end-expiratory pressure (PEEP)".)

Low levels of PEEP or zero PEEP may be indicated in patients with barotrauma or those who have a prolonged air leak from pneumothorax (provided PEEP is not indicated for severe ARDS). (See "Diagnosis, management, and prevention of pulmonary barotrauma during invasive mechanical ventilation in adults", section on 'Ventilator management' and "Management of persistent air leaks in patients on mechanical ventilation", section on 'Ventilator management'.)

Data discussing the effect of PEEP in patients who are mechanically ventilated intraoperatively and postoperatively and relative contraindications to PEEP are discussed separately. (See "Positive end-expiratory pressure (PEEP)", section on 'Intraoperative patients' and "Positive end-expiratory pressure (PEEP)", section on 'Postoperative patients' and "Positive end-expiratory pressure (PEEP)", section on 'Contraindications'.)

Fraction of inspired oxygen — Oxygenation goals should be individualized and hyperoxia should be avoided. In general, the following guidelines are reasonable:

●For most critically ill patients, the lowest possible FiO₂ necessary to meet oxygenation goals should be used, ideally targeting a peripheral arterial saturation (SpO₂) between 90 and 96 percent, if feasible [8]. This decreases the likelihood that adverse consequences of supplemental oxygen will develop, such as absorption atelectasis, accentuation of hypercapnia, airway injury, and parenchymal injury (see "Adverse effects of supplemental oxygen") while simultaneously avoiding dangerously low saturations. Maximum limits above which oxygen toxicity is certain are unclear and such limits have traditionally not been set. However, cumulating evidence suggests that for most acutely ill medical patients, SpO₂ >96 percent may not be necessary unless indications for high FiO₂ are present (eg, carbon monoxide toxicity, cluster headaches, sickle cell crisis, pneumothorax, pregnancy, air embolism). Similarly, while the safest lower limit has not been adequately studied, we typically use a target of approximately 90 percent, bearing in mind that most oximeters have a standard deviation of +/-2 when compared with measured saturations as well as other limitations. (See "Pulse oximetry".)

●For patients with hypercapnic hypoxemic respiratory failure, the limit of acceptable oxygenation may be lower. For example, an arterial oxygen tension (PaO₂) of 55 mmHg and a SpO₂ of 88 percent may be acceptable in patients with hypercapnia from chronic obstructive pulmonary disease (COPD). Further details are provided separately. (See "The evaluation, diagnosis, and treatment of the adult patient with acute hypercapnic respiratory failure", section on 'Titration of oxygen' and "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults", section on 'Inhaled pulmonary vasodilators' and "Management of refractory chronic obstructive pulmonary disease", section on 'Oxygen' and "Management of refractory chronic obstructive pulmonary disease", section on 'Nocturnal noninvasive ventilation'.)

●Oxygenation goals in patients with acute stroke or myocardial infarction are provided separately. (See "Initial assessment and management of acute stroke", section on 'Airway, breathing, and circulation' and "Overview of the acute management of ST-elevation myocardial infarction", section on 'Therapies of unclear benefit'.)

Several trials that included mostly critically ill patients agree that hyperoxia should be avoided. However, the definition of what constitutes hyperoxia varied among studies and data were derived from a heterogeneous population (eg, patients who were spontaneously breathing as well as from patients who were mechanically ventilated) [9-15]. In a retrospective single-center study of 2133 patients with 33,310 arterial blood gas measurements obtained during mechanical ventilation, any exposure to severe hyperoxemia (PaO₂ >200 mmHg) was associated with increased risk of 30-day mortality (odds ratio 1.29, 95% CI 1.04-1.59) [15]. Both the dose and duration of PaO₂ > 200 mmHg was associated with increased mortality.

Several meta-analyses have reported conflicting effects of a conservative oxygenation strategy compared with a liberal strategy:

●Meta-analysis in favor of an increase in mortality associated with a liberal strategy [10,11]:

•A meta-analysis of 25 randomized trials that included eight trials of patients who were mechanically ventilated compared a conservative oxygen strategy (FiO₂ 0.21; range 0.21-0.5) with a liberal oxygen strategy (median FiO₂ 0.52; range 0.28 to 1.0 for a median duration eight hours) [10]. Over 16,000 patients were studied including patients with sepsis, critical illness, stroke, trauma, myocardial infarction, or cardiac arrest, and patients who had emergency surgery. Patients in the liberal oxygen arm of the trials included in this analysis had a median baseline SpO₂ of 96 percent and the analysis excluded patients with hypoxemia (at baseline or on admission) as well as patients with chronic respiratory disease and extracorporeal membrane oxygenation (ECMO). A liberal oxygen strategy was associated with a small but increased hospital mortality (relative risk [RR] 1.21, 95% CI 1.03-1.43) and mortality at 30 days (RR 1.14, 95% CI 1.01-1.29). Importantly, as SpO₂ increased in the liberal strategy group, mortality also increased, indicating a dose-response relationship. There were no differences in morbidity outcomes including hospital acquired pneumonia and disability, or hospital length of stay.

●Meta-analyses demonstrating no appreciable difference in mortality associated with a liberal strategy [16-18]:

•In a network meta-analysis of eight trials of critically ill patients totaling 2532 patients requiring mechanical ventilation, there was no difference on 30-day mortality between patients who were treated with a conservative (PaO₂ 55 to 90 mmHg) versus liberal (PaO₂ >150 mmHg) oxygenation strategy (RR 0.89, 95% CI 0.61-1.30; low quality of evidence) [16]. However, many of the included studies were at high risk of bias, reducing confidence in the results.

•A meta-analysis of 17 trials of patients admitted to the ICU reported a marginal increase in the risk of mortality that was uncertain in patients treated with higher FiO₂ compared with patients treated with FiO₂ (RR 1.01, 95% CI 0.96-1.06) [18]. However, many of the included studies were at high risk of bias such that firm conclusions could not be made.

•Similarly, another meta-analysis of 50 randomized clinical trials that included 21,014 patients reported no difference in mortality rates between higher and lower oxygenation strategies in trials at overall low risk of bias (except for blinding; RR 0.98, 95% CI 0.89-1.09) [17].

Several randomized trials have also reported conflicting effects of a conservative oxygenation strategy compared with a liberal strategy:

●Clinical trials in favor of an increase in mortality associated with a liberal strategy [12,13]:

•One single center trial (included in the above meta-analysis) of 434 critically ill patients ventilated for respiratory failure due to a variety of etiologies randomized patients to a conservative oxygen therapy strategy (maintain a PaO₂ between 70 and 100 mmHg or SpO₂ between 94 and 98 percent) or a conventional oxygen strategy (PaO₂ up to 150 mmHg or SpO₂ between 97 and 100 percent) [12]. The conservative approach resulted in a lower mortality (12 versus 20 percent) and fewer episodes of shock (4 versus 11 percent), liver failure (2 versus 6 percent) and bacteremia (5 versus 10 percent). However, this trial should be interpreted with caution since it was terminated early for poor enrollment and used a modified intention to treat analysis. There were also baseline imbalances between the groups, and the study excluded patients with moderate to severe ARDS and patients with exacerbations of obstructive lung disease. In addition, the number of outcome events were small, which may have overestimated the benefit of the conservative strategy.

•In a randomized, open-label trial of 442 mechanically ventilated patients with septic shock, setting the FiO₂ to 1 (hyperoxia) for the first 24 hours was associated with a significant increase in ICU-acquired weakness (11 versus 6 percent) as well as a nonsignificant increase in mortality (43 versus 35 percent), compared with patients in whom the FiO₂ was set to a target SpO₂ of 88 percent [13]. However, the results should be interpreted with caution since some patients in this study may have also received hypertonic saline as part of their fluid resuscitation regimen.

●Data demonstrating no appreciable difference in mortality associated with a liberal strategy [14,18-22]:

•A multicenter randomized trial conducted in New Zealand and Australia (ICU-ROX) found no benefit from a conservative oxygen strategy. In this trial, no difference in the number of ventilator-free days or mortality was reported in 1000 ICU patients who were mechanically ventilated and treated with a conservative oxygen strategy (SpO₂ 90 to 97 percent) compared with usual oxygen strategy (SpO₂ >90 percent) [14]. However, the confidence limits were wide suggesting some patients may benefit. In addition, the target SpO₂ in the usual oxygen strategy group in this trial was considerably "less liberal" compared with other trials of conservative oxygen strategies, which may have limited the ability of the trial to detect a difference between the groups.

•Similarly, in a randomized trial of 205 patients with acute respiratory distress syndrome (ARDS) (LOCO₂), no difference in 28 day survival was reported between patients treated with a conservative strategy (target PaO₂ 55 to 70 mmHg SpO₂ 88 to 90 percent) and patients treated with a liberal strategy target PaO₂ 90 to 105 mmHg SpO₂ ≥96 percent (34 versus 27 percent) [19]. However, this study was stopped early for concerns over safety and futility of the conservative oxygen strategy and should be interpreted with caution.

•In another randomized trial of 2928 adults with acute hypoxemic respiratory failure, 60 percent of whom were mechanically ventilated, 90 day mortality was no different between the group of patients in whom a PaO₂ of 60 mmHg was targeted (low oxygen target) compared with patients in whom a PaO₂ of 90 mmHg was targeted (higher oxygen target; mortality rate 43 versus 42 percent) [20]. In addition, no difference in the percentage of ventilator-free days or adverse effects were reported. The slightly higher than expected mortality found in this trial compared with other trials may relate to a sicker population of patients.

•In a cluster-randomized, cluster-crossover trial of 2541 patients, there was no difference in ventilator-free days or 28-day mortality among the three different oxygen saturation target groups: low (88 to 92 percent), intermediate (92 to 96 percent) and high (96 to 100 percent) [21].

Flow rate and pattern — Inspiratory flow rates vary depending upon the mode of ventilation:

●During volume-limited ventilation, the rate of flow of air into the lung during inspiration is predetermined (ie, an inspiratory flow rate is set). An initial inspiratory flow rate is typically set between 60 and 70 L/minute (targeting an I:E ratio of 1:2 to 1:3). In most patients, these settings are sufficient to overcome pulmonary and ventilator impedance [23,24] and, therefore, minimize the work of breathing.

An insufficient inspiratory flow rate is characterized by dyspnea (from increased work of breathing), spuriously low inspiratory pressures, and scalloping of the inspiratory pressure tracing (figure 1) [25].

The need for a higher inspiratory flow rate is particularly common among patients who have obstructive airways disease with acute respiratory acidosis (eg, 60 to 75 L/minute). In such patients, a higher inspiratory flow rate shortens inspiratory time and increases expiratory time (ie, decreases the I:E ratio). These alterations increase carbon dioxide elimination and improve respiratory acidosis while also decreasing the likelihood of dynamic hyperinflation (auto-PEEP) and reducing work of breathing (figure 2) [25]. Although increased inspiratory flow rates can increase the peak airway pressure [26], the decreased inspiratory time may actually lower the mean airway pressure, which, in turn, can decrease oxygenation; in most cases reduced oxygenation is marginal and can be offset by increasing the FiO₂ if necessary. Inspiratory flow rates >75 L/minute are not typically needed and may be harmful. (See "Invasive mechanical ventilation in acute respiratory failure complicating chronic obstructive pulmonary disease", section on 'Dynamic hyperinflation'.)

●During pressure-limited ventilation, the inspiratory flow rate is not set but is determined by the inspiratory pressure limit, the inspiratory time, as well as the compliance/resistance of the respiratory system and patient effort. Thus, unlike volume-limited ventilation, the inspiratory flow is variable.

Mechanical ventilators can deliver several inspiratory flow patterns, including a square wave (constant flow), a ramp wave (decelerating flow), and a sinusoidal wave (waveform 1). The ramp wave is typically used since it may distribute ventilation more evenly than other patterns of flow, particularly when airway obstruction is present [27]. This decreases the peak airway pressure, physiologic dead space, and PaCO₂, while leaving oxygenation unaltered [28]. The effects of the different flow patterns on potential complications of mechanical ventilation (eg, hemodynamic impairment, pulmonary barotrauma, ventilator-associated lung injury) are unpredictable.

Trigger sensitivity — Breaths can be triggered by a timer (ventilator-initiated breaths) or by patient effort (patient-initiated breaths). Breaths that are initiated by a timer occur at the set respiratory rate and a trigger sensitivity does not need to be set, while those that are initiated by patient effort occur when the patient causes sufficient change in either the pressure (pressure-triggering) or flow (flow-triggering) in the circuit, in which case a trigger sensitivity needs to be set. Flow-triggering is often preferred because it is associated with less inspiratory effort in several modes of ventilation, even though the effect is usually small.

●When flow-triggering is used, a continuous flow of gas through the ventilator circuit is monitored by the ventilator. A ventilator-delivered breath is initiated when the return flow is less than the delivered flow, which occurs a consequence of the patient's effort to initiate a breath. During flow-triggering, the trigger sensitivity is usually set at 2 L/minute, which means that ventilator-assisted breaths will be triggered once the patient's inspiratory effort generates a flow of 2 L/minute (figure 3). Flow-triggering has been shown to decrease inspiratory work during continuous positive airway pressure and the spontaneous breaths of synchronized intermittent mandatory ventilation [29-31]. These studies that found that overall inspiratory effort was lower with flow-triggering than pressure-triggering during intermittent mandatory ventilation (IMV) [31,32] and PSV [33], and unaltered during assist control ventilation (ie, volume-limited or pressure limited assist control) [33].

●When pressure-triggering is used, a ventilator-delivered breath is initiated if the demand valve senses a negative airway pressure deflection generated by the patient trying to initiate a breath that is greater than the trigger sensitivity. A trigger sensitivity of -1 to -3 cm H₂O is typically set, which means that ventilator-assisted breaths will be triggered when the alveolar pressure decreases to 1 to 3 cm H₂O below atmospheric pressure. Pressure-triggering can be used with the assist control or synchronized intermittent mandatory ventilation modes of mechanical ventilation. Auto-PEEP (intrinsic positive end-expiratory pressure) interferes with pressure-triggering and this setting should not be used if auto-PEEP is suspected. The mechanism by which auto-PEEP interferes with pressure-triggering is discussed separately. (See "Positive end-expiratory pressure (PEEP)", section on 'Potential sequelae'.)

The trigger sensitivity should allow the patient to trigger the ventilator easily. A trigger sensitivity that is too sensitive may cause a breath to be delivered in response to patient movement or subtle pressure deflections caused by water moving within the ventilator tubing. In contrast, a trigger sensitivity that is not sensitive enough increases patient effort. This may cause a prolonged period between the initial effort and the ventilator breath or patient-ventilator asynchrony. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Treat dyssynchrony'.)

Intermittent sigh — Sighs are not routinely performed during mechanical ventilation. Moreover, not all ventilators can incorporate the sigh maneuver.

The term sigh refers to a deep breath, normally taken once every few minutes that is thought to maintain lung volume, avoid atelectasis, and improve lung compliance. The sigh, during mechanical ventilation, was popular in the past but fell out of favor when lung protective ventilation was found to be beneficial and high volumes shown to induce lung trauma. However, newer data suggest that the sigh maneuver, when limited and low in volume, may be safer than the more aggressive sigh strategies used in the past [34,35]. As an example, among 524 ventilated trauma patients who were at risk of ARDS, sigh volumes producing a plateau pressure of 35 cm H₂O delivered once every six minutes did not increase the number of ventilator-free days compared with those receiving usual care [35]. There was, however, a nonsignificant improvement in 28-day mortality. While there were few adverse events, hypotension attributable to sighs was seen in 2 percent and appeared to be associated with a vasopressor requirement. Further data are needed before sighs can be routinely applied during mechanical ventilation.

SPECIFIC POPULATIONS — Select populations need specific considerations and are discussed separately:

●Acute respiratory distress syndrome (ARDS) (see "Acute respiratory distress syndrome: Ventilator management strategies for adults")

●Respiratory failure complicating chronic obstructive pulmonary disease or asthma (see "Invasive mechanical ventilation in acute respiratory failure complicating chronic obstructive pulmonary disease" and "Invasive mechanical ventilation in adults with acute exacerbations of asthma")

●Traumatic brain injury (see "Management of acute moderate and severe traumatic brain injury", section on 'Ventilation')

●Patients with an air leak (see "Management of persistent air leaks in patients on mechanical ventilation", section on 'Ventilator management' and "Diagnosis, management, and prevention of pulmonary barotrauma during invasive mechanical ventilation in adults", section on 'Ventilator management')

●Patients undergoing anesthesia (see "Mechanical ventilation during anesthesia in adults")

●Patients undergoing bariatric surgery (see "Intensive care unit management of patients with obesity", section on 'Mechanical ventilation')

FOLLOW-UP — While most patients tolerate volume-limited or pressure-limited assist-controlled ventilation, some patients do not tolerate it as evidenced by dyssynchrony, high airway pressures, or hypoxemia. Thus, patients should be followed clinically and bedside ventilator waveforms should be examined on a regular basis. In such cases, an adjustment of settings or a change in the mode is often necessary. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Patients who are not improving or deteriorating'.)

GENERAL CARE — The supportive care of patients who are mechanically ventilated include the following and are discussed separately:

●Sedation, analgesia, and delirium management (see "Sedative-analgesia in ventilated adults: Management strategies, agent selection, monitoring, and withdrawal" and "Sedative-analgesia in ventilated adults: Medication properties, dose regimens, and adverse effects" and "Pain control in the critically ill adult patient" and "Neuromuscular blocking agents in critically ill patients: Use, agent selection, administration, and adverse effects" and "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults", section on 'Sedation' and "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults", section on 'Paralysis (neuromuscular blockade)')

●Hemodynamic monitoring (see "Pulmonary artery catheterization: Indications, contraindications, and complications in adults" and "Novel tools for hemodynamic monitoring in critically ill patients with shock")

●Nutritional support (see "Nutrition support in intubated critically ill adult patients: Initial evaluation and prescription")

●Glucose control (see "Glycemic control in critically ill adult and pediatric patients")

●Measures to prevent ventilator-associated pneumonia (see "Risk factors and prevention of hospital-acquired and ventilator-associated pneumonia in adults")

●Venous thromboembolism prophylaxis (see "Prevention of venous thromboembolic disease in acutely ill hospitalized medical adults" and "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement" and "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients")

●Gastrointestinal prophylaxis (see "Stress ulcers in the intensive care unit: Diagnosis, management, and prevention", section on 'Prophylaxis')

●Venous or arterial access (see "Central venous access in adults: General principles" and "Intra-arterial catheterization for invasive monitoring: Indications, insertion techniques, and interpretation")

●Temperature management (see "Fever in the intensive care unit" and "Management of acute moderate and severe traumatic brain injury", section on 'Temperature management')

SUMMARY AND RECOMMENDATIONS

●Definition – Invasive mechanical ventilation is defined as the delivery of positive pressure to the lungs via an endotracheal or tracheostomy tube. (See 'Definition' above.)

●Indications – Invasive mechanical ventilation is most often used to fully or partially replace the functions of spontaneous breathing by performing the work of breathing and gas exchange (table 1 and table 2 and table 3). Invasive mechanical ventilation may also be useful in those who require airway protection (eg, depressed mental status from an overdose, patients with variceal bleeding). (See 'Indications' above.)

●Selecting initial mode – There is no universal initial mode of ventilation that is ideal for all patients (table 4). However, common initial modes, which are suitable for most patients, include the following:

•Volume-limited assist control ventilation (including pressure-regulated volume-control [PRVC])

•Pressure-limited assist control ventilation

•Synchronized intermittent mandatory ventilation with pressure support ventilation (SIMV-PSV)

Factors that influence the initial mode chosen include the level of ventilator support needed, reason for mechanical ventilation, presence of airflow limitation, presence of an air leak, elevated intracranial pressure, need for heavy sedation/paralysis, and institutional or clinician preferences. (See 'Selecting an initial mode' above.)

●Initial settings – Once the mode of ventilation has been selected, we typically use the following initial ventilator settings (table 5) (see 'Settings' above):

•Volume-limited assist control ventilation – For patients in whom volume-limited assist control ventilation is chosen, typical initial settings include the following (see 'Volume-limited assist control ventilation' above and "Modes of mechanical ventilation", section on 'AC'):

-A tidal volume of 6 mL per kg of predicted body weight (PBW; range is 4 to 8 mL/kg PBW (table 6 and table 7 and table 8) (calculator 1 and calculator 2))

-A respiratory rate of 12 to 16 breaths per minute (higher rates up to 35 breaths/minute may be needed for low tidal volume ventilation in acute respiratory distress syndrome [ARDS])

-An applied positive end-expiratory pressure (PEEP) of 5 cm H₂O (PEEP may be adjusted to the fraction of inspiratory oxygen [FiO₂] in ARDS (table 6))

-An FiO₂ sufficient to meet oxygenation goals (ideally 90 to 96 percent unless indications for higher peripheral oxygen saturation [SpO₂] are present)

-An inspiratory flow rate between 40 and 60 L/minute (targeting an inspiratory to expiratory [I:E] ratio of 1:2 to 1:3)

-A trigger sensitivity -2 L/minute (when flow-triggering is used) or -1 to -2 cm H₂O (when pressure-triggering is used)

If PRVC is used, settings are similar, except instead of setting the inspiratory flow, inspiratory time is set to deliver an I:E ratio of 1:2 to 1:3.

•Pressure-limited assist control ventilation – For patients in whom pressure-limited assist control ventilation is chosen, we typically set the following (see 'Pressure-limited assist control ventilation' above and "Modes of mechanical ventilation", section on 'Pressure-limited ventilation'):

-The inspiratory pressure level is set targeting an approximate tidal volume (eg, typically between 12 and 25 cm H₂O to target 4 to 8 mL/kg predicted body weight [PBW])

-The inspiratory time is set to deliver an I:E ratio of 1:2 to 1:3

-The FiO₂, ventilator rate, applied PEEP, and trigger sensitivity are similar to those of volume-limited assist control ventilation

The initial inspiratory pressure required can vary depending upon the patient's lung compliance and airway resistance, as well as tubing resistance, but in general, acceptable target tidal volumes may be reached with inspiratory pressure levels between 12 and 25 cm H₂O.

•SIMV-PSV – For patients in whom SIMV-PSV is chosen, we typically use similar settings to volume-limited assist control ventilation with the addition of pressure support for spontaneous breaths taken by the patient above the set rate (eg, 5 to 10 cm H₂O). (See 'Synchronized intermittent mandatory ventilation with pressure support ventilation (SIMV-PSV)' above and "Modes of mechanical ventilation", section on 'SIMV'.)

●Special populations – Select populations who deserve special attention include patients with the following:

•ARDS (see "Acute respiratory distress syndrome: Ventilator management strategies for adults")

•Respiratory failure complicating airflow obstruction (see "Invasive mechanical ventilation in acute respiratory failure complicating chronic obstructive pulmonary disease" and "Invasive mechanical ventilation in adults with acute exacerbations of asthma")

•Traumatic brain injury (see "Management of acute moderate and severe traumatic brain injury", section on 'Ventilation')

•Patients with an air leak (see "Diagnosis, management, and prevention of pulmonary barotrauma during invasive mechanical ventilation in adults", section on 'Ventilator management')

•Patients undergoing anesthesia (see "Mechanical ventilation during anesthesia in adults")

•Patients undergoing bariatric surgery (see "Intensive care unit management of patients with obesity", section on 'Mechanical ventilation')

●Follow-up and general care – Following the initiation of invasive mechanical ventilation, we perform the following routinely (see 'Follow-up' above and 'General care' above):

•Patients should be followed clinically daily using history, examination, laboratory, microbiologic, and radiologic data as needed.

•Bedside ventilator waveforms should be examined on a regular basis, so that the mode and settings can be adjusted according to the patient's needs (waveform 1).

•Several issues surrounding supportive care should also be addressed, including sedation and analgesia, hemodynamic monitoring, nutrition, glucose control, prevention measures for ventilator-associated pneumonia, venous thromboembolism and gastric ulceration, venous and arterial access, and temperature management. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults".)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Anthony Courey, MD, who contributed to earlier versions of this topic review.