Invasive mechanical ventilation in acute respiratory failure complicating chronic obstructive pulmonary disease

INTRODUCTION — Invasive mechanical ventilation is a means of life support typically reserved as a last option for acute respiratory failure in chronic obstructive pulmonary disease (COPD). Mechanical ventilation in this population is specifically associated with complications including dynamic hyperinflation and barotrauma that can lead to cardiovascular collapse and death.  

Details of invasive mechanical ventilation for acute respiratory failure in patients with COPD are discussed in this topic. The use of noninvasive positive pressure ventilation and other aspects of the management of acute respiratory failure in COPD are reviewed separately. (See "Noninvasive ventilation in adults with acute respiratory failure: Benefits and contraindications" and "COPD exacerbations: Management".)

INDICATIONS — The indications for intubating patients with chronic obstructive pulmonary disease (COPD) are similar to those for all patients with acute respiratory failure (eg, life-threatening respiratory distress, failure of oxygenation and ventilation). The decision to institute invasive mechanical ventilation in this population is typically based upon a constellation of clinical signs and symptoms in the context of the patient's preferences for life support. While in the past, patients were traditionally intubated late in the course of their acute illness (eg, respiratory arrest) clinical practice now supports earlier intubation. (See "The decision to intubate" and "Rapid sequence intubation in adults for emergency medicine and critical care".)  

The majority of patients with acute respiratory failure due to acute exacerbations of COPD (AECOPD) should undergo a trial of noninvasive mechanical ventilation (NIV) because intubation can be successfully avoided and mortality reduced with this intervention [1-7]. Exceptions include patients with contraindications to NIV, failure of NIV, and severe respiratory distress. These are summarized below. (See "Noninvasive ventilation in adults with acute respiratory failure: Benefits and contraindications".)

This initial approach can take the form of noninvasive positive pressure ventilation (NIPPV) via nasal or full facemask, but can also be approached with oxygen delivered through high flow nasal cannula (HFNC). There are now data in patients with hypoxic respiratory failure that support the use of HFNC as a means of avoiding the need for intubation [8]. In fact, preliminary data suggest that HFNC may be more effective than NIPPV in avoiding the need for intubation, and is potentially associated with fewer long-term complications than NIPPV [8]. (See "Continuous oxygen delivery systems for the acute care of infants, children, and adults", section on 'Nasal cannula'.)

Life threatening respiratory failure — Immediate intubation and mechanical ventilation should be considered in those with features of life-threatening respiratory failure. Examples of such patients include those with severe respiratory distress who are not suitable for NIV as well as those with agonal breathing, cardiopulmonary arrest, and concurrent severe hemodynamic instability [9,10].

A bedside scoring system, BAP-65, that uses signs of respiratory distress, along with other risk factors, has been found to predict the need for mechanical ventilation in patients with acute exacerbations of COPD [11,12]. In one study, patients with four risk factors (elevated BUN, altered mental status, pulse >109 beats/min, age >65 years) were more likely to require mechanical ventilation compared to those with none of these risks (55 versus 2 percent).  

Failure of noninvasive ventilation — Several studies in patients with AECOPD have shown that deteriorating or unaltered gas exchange despite NIV predicts the need for invasive mechanical ventilation [6,13,14]. In one multicenter prospective trial of 1033 patients with AECOPD, patients with the following features at initial presentation had a greater than 70 percent chance of failing NIV and requiring mechanical ventilation [6]: Glasgow Coma Score <11, acute physiology and chronic health evaluation (APACHE) II score ≥29, respiratory rate ≥30, and pH <7.25. After two hours of NIV, a pH <7.25 further increased the likelihood of need for intubation from 70 to 90 percent. These data together with clinical experience support the need for close monitoring in an intensive care unit for patients with AECOPD initially treated with NIV. Interval evaluation at two hours should be considered so that timely escalation in the level of care from noninvasive to invasive ventilation can proceed.

Trends in practice and prognosis appear to favor use of NIV over IMV in COPD patients hospitalized for respiratory failure. Between 2001 and 2011, initial use of NIV over IMV has steadily increased by 15 percent annually while initial IMV use has steadily declined by 3.2 percent annually [7]. A cross-sectional analysis of 77,576 patients hospitalized in 386 hospitals between 2009 and 2011 demonstrated considerable hospital variation in the use of NIV as the initial mechanical ventilation strategy, with the majority turning to NIV before IMV [15]. This study demonstrated that higher rates of risk-standardized NIV use were associated with reduced hospital costs, shorter length of stay, and lower adjusted mortality compared with IMV [15]. A separate cohort of 25,628 patients with AECOPD throughout 420 hospitals found that 70 percent were initially treated with NIV. Compared with those initially on IMV, use of NIV was associated with lower rates of hospital acquired pneumonia, lower hospital costs, and a reduced length of stay after propensity matching.

While NIV has not been compared with high flow oxygen delivered via nasal cannula (HFNC), some experts are increasingly using HFNC in the setting of AECOPD. Whether this strategy definitively reduces intubation rates in patients with AECOPD is yet to be determined. Further details regarding HFNC are discussed separately. (See "Heated and humidified high-flow nasal oxygen in adults: Practical considerations and potential applications".)

Arterial blood gas abnormalities — In patients with COPD, there are no absolute values on arterial blood gas (ABG) that determine when mechanical ventilation should be instituted. Typically, we consider intubation when the following gas exchange abnormalities are present, particularly after a trial of NIV:  

●Hypoxemia that has not corrected with supplemental oxygen (failure to oxygenate)

●Severe respiratory acidosis, unresponsive to therapy and/or NIV (failure to ventilate)

Many patients with COPD who present with acute respiratory failure have chronic hypoxia and/or hypercapnia due to their underlying disease or other comorbid conditions (eg, obesity hypoventilation). The interpretation of an ABG in this population should always be done in the context of prior baseline values when available, and the knowledge of influencing comorbidities. Additionally, paying attention to the presence of mixed acid-base disorders and calculating the predicted pH for acute and chronic acidosis should be done to avoid unnecessary intubation for chronic conditions. For example, a COPD patient with an arterial carbon dioxide tension (PaCO₂) of 100 mmHg may appear alarming to the clinician at the bedside. However, if the pH is near normal (eg, 7.32) or there is evidence of a prior baseline PaCO₂ close to the measured value (eg, 90 mmHg), this suggests that the degree of acute hypercapnia is likely minor and intubation may potentially still be avoided. (See "Arterial blood gases" and "Simple and mixed acid-base disorders".)

GOALS OF VENTILATOR MANAGEMENT — The primary goals of invasive mechanical ventilation in patients with chronic obstructive pulmonary disease (COPD) are the following [16]:

●Correct derangements in oxygenation and ventilation

●Reduce the work of breathing

●Prevent dynamic hyperinflation (DHI)

The prevention of DHI is especially important in this population because DHI can lead to significant complications such as barotrauma, which can prolong the course of mechanical ventilation, and result in cardiovascular collapse or death. (See "Dynamic hyperinflation in patients with COPD" and 'Dynamic hyperinflation' below and "Clinical and physiologic complications of mechanical ventilation: Overview", section on 'Auto-PEEP'.)

VENTILATOR MODES

Commonly used modes — A number of modes of mechanical ventilation can be used to ventilate patients with respiratory failure and chronic obstructive pulmonary disease (COPD). Volume or combined volume and pressure control modes of ventilation are commonly used. For this section, each mode of ventilation will be classified according to the most recently promoted taxonomy, based primarily upon its control variable, breath sequence, and targeting scheme [17]. It's important to note, however, that there are literally dozens of combinations of these three parameters, and despite sharing a similar combination of these parameters, such modes can carry different proprietary labels from one manufacturer to the next, as in the case of "CMV with autoflow" and "Pressure-regulated volume control." At the same time, modes that appear similar to one another across manufacturers can vary by the sake of subtle differences in their targeting schemes. For simplicity, this chapter will refer only to the basic taxonomy label dictated by the control variable and breath sequence. Although the optimal mode is unknown, we typically use the following volume-targeted modes: volume control continuous mandatory ventilation (VC-CMV), sometimes referred to as "assist-control," and synchronized intermittent mandatory ventilation (SIMV) or SIMV with pressure support ventilation (SIMV/PSV). Switching modes is not uncommon during the course of a patient's illness particularly when complications of mechanical ventilation, such as acute respiratory distress syndrome (ARDS) or barotrauma occur.

The advantages and disadvantages of the modes of ventilation typically used in patients with acute respiratory failure complicating COPD are reviewed in this section. The details of these modes as well as other modes of mechanical ventilation are reviewed separately. (See "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit" and "Modes of mechanical ventilation".)

Volume-control (VC) and volume "targeted," pressure control (PC) modes — In VC modes of ventilation, volume and flow are preset, such that ventilator-assisted breaths are triggered by either the ventilator (time) or the patient, and a set volume is delivered. For some modes, such as "CMV with autoflow," although a target volume is "preset," inspiratory pressure varies with breath to breath to achieve an average tidal volume using an adaptive targeting scheme, and is thus technically a pressure control (PC) mode (PC-CMV).

Volume control-continuous mandatory ventilation (VC-CMV) — VC-CMV is a volume control mode that delivers a continuous mandatory breath sequence, with no allowance for spontaneous breaths. This translates to a preprogrammed number of mandatory breaths, triggered by time, with the allowance of additional patient-triggered mandatory breaths in between. By definition of being a volume control mode, both the tidal volume and flow are preset [17].When pressure is regulated breath to breath to achieve an average target volume during CMV, the mode is technically pressure control (PC-CMV). Both modes share similar objectives and drawbacks.

The key benefit from PC-CMV and VC-CMV modes is the ability to guarantee a predetermined minimum minute ventilation (set tidal volume multiplied by the set ventilator rate), which is generally high in patients with COPD due to abnormally increased physiologic dead space. Both modes can more easily meet this demand. The greatest disadvantage of these modes in patients with COPD is overventilation when the patient is tachypneic and frequently triggering breaths, all of which are mandatory in volume. Overventilation can lead to severe respiratory alkalosis and dynamic hyperinflation (DHI), which can lead to barotrauma, cardiovascular collapse, and even death. Among all modes, the work of breathing is lowest in patients receiving PC-CMV and VC-CMV [18,19]. However, the work of breathing on this mode can still be excessively high if the patient is not adequately sedated, is dyssynchronous with the ventilator, or has excessively high inspiratory flow demands [20,21]. Thus, in patients with COPD, careful selection of settings (to avoid excessively high minute ventilation), judicious use of sedatives (to decrease respiratory drive and work), and aggressive bedside monitoring (for DHI) are necessary for the success with these modes. (See 'Dynamic hyperinflation' below and "Modes of mechanical ventilation", section on 'Volume-limited ventilation'.)

Intermittent mandatory ventilation (SIMV) — An IMV breathing pattern can be delivered under volume or pressure control conditions (VC-IMV or PC-IMV, respectively) [17]. Most IMV is synchronized to patient effort (SIMV) using computer-controlled synchronization windows to better accommodate patient demands. Like VC-CMV, VC-IMV is programmed to deliver a predetermined minimum minute ventilation, using a preset rate of time-triggered mandatory breaths. However, unlike VC-CMV, VC-IMV is accompanied by windows of time that allow for additional spontaneous breaths, which are both triggered and cycled (terminated) by the patient. Each of these spontaneous breaths can either be entirely unsupported (VC-IMV) or partially assisted by pressure support (VC-IMVs,s) [17].

VC-IMV and PC-IMV, with or without pressure support, are alternative modes for use in patients with COPD who generate excessively high minute ventilation on CMV. Because not every breath is mandatory and preprogrammed in size, switching from CMV to IMV can reduce the risk of DHI and its attendant complications. Compared with CMV, this mode has been shown to be associated with greater work of breathing, which may offset some of the perceived benefits [21,22]. Notably, when patients have stabilized and liberation from mechanical ventilation is being planned, using an IMV mode should be discouraged, as this mode is associated with slower weaning from mechanical ventilation when compared to spontaneous breathing trials [23]. (See "Modes of mechanical ventilation", section on 'Volume-limited ventilation'.)

Pressure control-continuous spontaneous ventilation (PC-CSV) — In this mode, commonly referred to as "pressure-support" ventilation, the breathing pattern is continuously spontaneous (CSV) with each breath being supported by a preset level of pressure (ie, pressure control [PC]). As a CSV mode, all breaths are patient triggered and cycled (terminated), and the volume of each spontaneous breath is variable, but augmented with PC support.

Unlike VC-CMV and VC-IMV, minute ventilation is determined by patient effort and respiratory mechanics, and consequently, a predetermined minute ventilation is not guaranteed. Work of breathing is also variable with PC-CSV. Depending on the level of support provided, work of breathing may be greater, equal to, or even less than that observed on CMV modes in patients ventilated predominantly for conditions other than COPD [24]. When compared with unsupported breaths ("pressure support" = 0) and IMV alone, incremental levels of pressure support in the low range (ie, between 0 and 60 percent of patient effort) decrease patient effort and work of breathing. However, the incremental benefit diminishes at higher levels of pressure support (>60 percent of patient effort) (figure 1) [18,19]. Consistent with this observation, patients with COPD can become more dyssynchronous with the ventilator when higher levels of support are used (waveform 1) [25].

PPC-CSV is infrequently used to ventilate patients with COPD. This is, in part, because a predetermined minute volume cannot be set, and reduced work of breathing is not guaranteed with this mode, both of which are desirable in COPD [18,19]. In addition, minute ventilation may be insufficient during PC-CSV in patients with COPD because airway resistance is high. Since each breath is spontaneous, pressure support typically cycles off below a threshold of inspiratory effort (flow), and since high airway resistance results in decreased airflow during inspiration, this can cause inspiration to terminate prematurely before the optimal tidal volume has been delivered [9,26]. PSV also does little to avoid auto-PEEP (the product of dynamic hyperinflation), which when left untreated, can lead to increased patient work and barotrauma [27]. (See "Modes of mechanical ventilation", section on 'Pressure support'.)

VENTILATOR SETTINGS

Initial ventilator settings and monitoring — The ideal ventilator settings in patients with acute respiratory failure complicating chronic obstructive pulmonary disease (COPD) are those that meet the goals of mechanical ventilation: normalizing oxygenation and ventilation, reducing the work of breathing, and avoiding dynamic hyperinflation. The settings will vary according to the cause of acute respiratory failure as well as the mode of mechanical ventilation selected, which is VC-CMV or PC-CMV. Although not absolute, we typically initiate mechanical ventilation with the following initial settings for each mode:

●VC/PC-CMV – For patients in whom CMV is chosen as the initial mode of ventilation, we typically use the following settings:

•Fraction of inspired oxygen (FiO₂) to maintain the oxygen saturation of hemoglobin (SO₂) above 92 percent.

•Tidal volume – 6 to 8 mL/kg (lower starting tidal volumes of 4 to 6 mL/kg may be preferred in COPD patients ventilated for ARDS)

•Ventilator rate – 10 to 16 breaths per minute.

•Minute ventilation – Taking into consideration the aforementioned tidal volumes and rates, target minute ventilation could range from 60 to 128 mL/kg, but one would typically match a slightly higher rate with a lower tidal volume. One must also take into consideration the patient's physiologic dead space, which will be higher in COPD, and the dead space to tidal volume ratio, which will increase with smaller tidal volumes.

•Positive end expiratory pressure (PEEP) – 5 to 10 cm H₂O.

•Inspiratory flow – 60 L/min.

•Trigger sensitivity – -1 to -2 cm H₂O when pressure triggering is used or 2 L/min when flow triggering is used.

●VC/PC-IMV – For patients in whom volume control, synchronized intermittent mandatory ventilation with pressure support (VC-IMVs) is chosen as the initial mode of ventilation, we typically use similar settings to those of VC-CMV listed above, with the addition of pressure support (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 the objective of using IMV support.

●PC-CSV – Should this mode be chosen, we generally increase the pressure support level until the patient's respiratory rate falls below 30 breaths per minute (no ventilator or time trigger exists in CSV). However, when using PC-CSV, achieving the optimal pressure support level can be difficult in COPD. This is because the expiratory effort ("fighting the ventilator"), and consequently the respiratory rate, of patients with COPD may actually increase when the level of support is increased [27]. There is some evidence, with noninvasive modes, that PC-CSV can be less effective than VC/PC-CMV modes in reducing work of breathing, but can be perceived (by the observing provider) as being "more comfortable" to patients, likely because of "spontaneous" breaths being triggered and cycled by the patient [28]. This suggests it may be reasonable to transition COPD patients over from VC/PC-CMV to PC-CSV (with sufficient pressure support to conserve work of breathing and keep respiratory rate less than 30) once these patients have had sufficient time to recover from the more severe phase of their exacerbation. The FiO₂, applied PEEP, inspiratory flow, and trigger sensitivity are chosen similarly to those of AC-CMV.

Due to abnormally high physiologic dead space, patients with COPD will often have higher than expected minute ventilation requirements and increased carbon dioxide retention, which typically improves over the patient's course [29,30]. The initial settings should be adjusted to satisfy the patient's minute ventilation requirement. Adjusting ventilator settings according to ABG results alone is problematic because these data provide no information regarding the work of breathing, which may be excessive, even when gas exchange is acceptable [21]. Subsequent adjustments to ventilator settings are best made at the bedside, according to the patient's response, respiratory effort, the presence of auto-PEEP, as well as arterial blood gas values (usually done within 30 minutes of any major adjustment).

Fraction of inspired oxygen — Based upon findings from a systematic review, in which it was found that supplemental oxygen delivered to patients with normal oxygen saturation was associated with increased mortality [31], and an expert consensus panel guideline for supplemental oxygen delivery in acutely hospitalized patients [32], we advocate that supplemental oxygen be held or down-titrated for a peripheral oxygen saturation (SpO₂) of greater than 96 percent. Because their recommendation for the lower limits of an SpO₂ threshold for starting supplemental oxygen therapy were largely informed by trials involving patients with stroke or myocardial infarction, the consensus panel was uncertain as to how to guide therapy for patients outside of this cohort. Nevertheless, given the nonlinear and steep nature of the hemoglobin dissociation curve below an SpO₂ of 90 percent, logic would dictate that supplemental oxygen be provided for COPD patients at an SpO₂ of 90 percent or below, and titrated to between 93 and 96 percent. These settings provide a safe cushion against dangerous desaturations in oxygenation, while at the same time avoiding any potential toxicity from high levels of oxygen [33]. (See "Adverse effects of supplemental oxygen" and "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Fraction of inspired oxygen'.)

Typically, it is safe to adjust the FiO₂ while continuously monitoring SpO₂ by pulse oximetry. However, the clinician should periodically check SpO₂ measured by arterial blood gas co-oximetry due to the inaccuracy of peripheral monitoring in select circumstances (eg, nail polish, African Americans). (See "Pulse oximetry".)

Hypoxemia is rarely refractory to supplemental oxygen in patients with exacerbations of COPD. However, this does not apply to patients with COPD who are ventilated for conditions characterized by high degrees of shunt physiology (eg, acute respiratory distress syndrome or patent foramen ovale). (See "Measures of oxygenation and mechanisms of hypoxemia" and "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults".)

Tidal volume — The optimal tidal volume for patients with acute respiratory failure complicating COPD is unknown and is largely determined by the etiology of respiratory failure. A low tidal volume strategy (eg, 4 to 6 mL/kg) is beneficial in patients with acute respiratory distress syndrome (ARDS) and is appropriate for COPD patients who are mechanically ventilated for ARDS. For patients with COPD who are ventilated due to other reasons, tidal volumes in the low range (eg, 6 to 8 mL/kg) may be tried as an initial strategy, although the benefit is less proven. The rationale for this approach is based upon the known high minute ventilation requirements in patients with COPD and the desire to avoid overdistension, dynamic hyperinflation, and lung injury in these patients, who are already at risk for these complications. (See "Acute respiratory distress syndrome: Ventilator management strategies for adults", section on 'Low tidal volume ventilation: Initial settings' and "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Tidal volume'.)

Ventilator rate — The optimal ventilator rate for patients with acute respiratory failure complicating COPD is not established and is largely determined by the patient's native respiratory rate. The ventilator rate (VR) refers to the number of breaths per minute that is set by the clinician on the ventilator (mandatory breaths). The spontaneous rate (SR) refers to the number of breaths per minute that a patient receives beyond the set ventilator rate (non-mandatory breaths). The patient's respiratory rate (RR) refers to the total number of breaths that the patient receives in one minute (RR = VR + SR). Although there is no desired upper or lower limit, we typically aim for RR <25 (ideally <20) and >8 to 10 breaths per minute. (See "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Ventilator rate'.)

●During ACV, as a general rule, the VR is often set four breaths per minute less than the RR (eg, VR is set at 16 when the patient's RR is 20 breaths per minute). However, the patient should be monitored closely because a particularly high respiratory rate will decrease the expiratory time. This can worsen dynamic hyperinflation and result in inverse ratio ventilation (ie, inspiratory time is longer than the expiratory time), which is not desirable in COPD.

●During SIMV, the initial VR is typically set to ensure at least 80 percent of the patient's total minute ventilation (RR multiplied by tidal volume) is delivered by the ventilator. For example, if the patient's minute ventilation is recorded by the ventilator at 10 L/min with a set tidal volume of 0.4 L, then the rate would be set at 20 breaths per minute (20 x 0.4 = 8 L).

●A ventilator rate is not set during PSV.

The target rate is frequently adjusted to target a total minute ventilation (tidal volume multiplied by rate) of 115 mL/kg. The set rate can be decreased as needed to mitigate dynamic hyperinflation. (See 'Dynamic hyperinflation' below.)

Applied PEEP — It is now recognized that applied PEEP may be beneficial in patients with expiratory airflow limitation (eg, COPD), despite past concerns that applied PEEP at any level could worsen dynamic hyperinflation (also known as auto-PEEP). Ideally, the application of extrinsic PEEP should slightly "undermatch" intrinsic PEEP so as to avoid worsening of auto-PEEP. In other words, extrinsic PEEP should never be more than intrinsic PEEP. In practice this is difficult to achieve because intrinsic PEEP is dynamic and fluctuates over time, often decreasing with therapy. Traditionally in the past, applied PEEP at levels no greater than 80 percent of intrinsic PEEP was advised, given the difficulty of actually measuring intrinsic PEEP in this population, together with the dynamic nature of auto-PEEP, and the harmful effects of too much applied PEEP. We now typically initially set PEEP at 5 to 10 cm H₂O and adjust accordingly depending on the patient's course. Alternatively, starting at 5 cm H₂O and increasing PEEP only if ineffective trigger efforts are observed can also be effective. Ineffective efforts are due to the patient having to overcome intrinsic PEEP before their effort is sensed by the ventilator. Ineffective efforts can often be alleviated with applied, extrinsic PEEP (provided it is not equal or greater than the intrinsic PEEP), which in turn may actually help to corroborate suspicions regarding the presence of auto-PEEP. For example, resolution of ineffective efforts with applied extrinsic PEEP can help confirm the presence of auto-PEEP and act as a guide to avoid the application of excessive extrinsic PEEP. (See "Positive end-expiratory pressure (PEEP)" and 'Prevention and treatment' below.)

Inspiratory flow rate — As a general rule, we suggest that the initial inspiratory flow rate be set at 60 L/min. The rate can then be adjusted (usually increased) if the patient appears to be dyssynchronous due to a high work of breathing (ie, higher inspiratory flow demand) or has dynamic hyperinflation. (See 'Dynamic hyperinflation' below and "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Flow rate and pattern'.)

A high inspiratory flow rate (60 to 100 L/min) is generally advised in mechanically ventilated patients with COPD, to decrease the work of breathing and reduce the risk of dynamic hyperinflation (DHI) [34-36]:

●The inspiratory work necessary to overcome pulmonary and ventilator resistance in COPD is markedly increased if the inspiratory flow is insufficient. This is because patients with COPD generally have an increased respiratory (and inspiratory flow) demand. Increasing the inspiratory flow rate will often satisfy this high demand and decrease the work of breathing [34,35,37].

●A high inspiratory flow rate is necessary to shorten inspiratory time and hence prolong expiratory time. A prolonged expiratory time is critical in patients with expiratory airflow limitation, by giving the patient time to fully exhale a delivered breath. As a result of complete or near complete exhalation, the likelihood of dynamic hyperinflation and consequent auto-PEEP decreases.

A high inspiratory flow rate is occasionally associated with an increase in the spontaneous respiratory rate, which has the potential to worsen dynamic hyperinflation [35,38]. However, the benefits of prolonged expiratory time appear to outweigh this risk. In one study of patients with COPD, increasing the inspiratory flow rate from 30 to 90 L/minute increased the respiratory rate from 16 to 20 breaths per minute (figure 2) [35]. However, despite an increase in respiratory rate, the net effect of increased inspiratory flow was to prolong expiratory time (2.1 to 2.3 seconds) and reduce auto-PEEP (7 to 6.4 cm H₂O).

Although the main purpose of increasing inspiratory flow is a prolongation of the expiratory time, in many cases, greater prolongation of the expiratory time can be achieved by suppressing the patient's intrinsic respiratory drive and decreasing the respiratory rate. (See 'Prevention and treatment' below.)

Trigger sensitivity — Ventilators can be triggered to deliver a breath by either a change in airway opening pressure (ie, pressure triggered) or flow (ie, flow triggered) (figure 3). Most modern ventilators today use flow triggering, integrated with a continuous low-level bias flow. We do not have a preference for pressure or flow triggering in patients receiving CMV. Inappropriate breaths may be initiated if the trigger setting is too sensitive, which can cause respiratory alkalosis. Conversely, the work of breathing may be increased if the trigger setting is not sensitive enough. (See "Overview of initiating invasive mechanical ventilation in adults in the intensive care unit", section on 'Trigger sensitivity'.)

●The trigger sensitivity is usually set at -1 to -2 cm H₂O when pressure triggering is used. This means that ventilator-assisted breaths will be triggered when the airway pressure decreases to 1 to 2 cm H₂O below atmospheric pressure.

●The trigger sensitivity is usually set at 2 L/min when flow triggering is used. This means that ventilator-assisted breaths will be triggered once the patient's inspiratory effort generates a flow of 2 L/min above the bias flow running through the circuit.

When using an IMV setting and relying heavily upon patient-triggered spontaneous breaths, it may be preferable (when given a choice) to favor flow triggering. In patients receiving IMV, total inspiratory effort is 30 to 40 percent less with flow triggering than pressure triggering [39,40]. In contrast, in patients receiving PC-CSV, inspiratory effort during triggering is only 10 percent less with flow triggering than pressure triggering [41]. However, this is of minimal clinical importance because the difference occurs during triggering only, which is a small part of the total inspiratory effort [19,41]. Inspiratory work does not appear to be impacted by the triggering method during ACV [41].

DYNAMIC HYPERINFLATION

Auto-PEEP — Dynamic hyperinflation (DHI) is characterized by increased levels of intrinsic positive end-expiratory pressure (PEEPi or "auto-PEEP"). The hyperinflation is progressive (dynamic) because air accumulates in the lung with each breath as a result of a failure to achieve complete exhalation before the onset of the next breath (figure 4). In patients with COPD who are intubated for respiratory failure, DHI can occur as a consequence of airflow obstruction due to bronchoconstriction, combined with a higher than normal minute ventilation (respiratory rate multiplied by tidal volume) delivered by the ventilator. DHI (spontaneous or ventilator-induced) creates elevated levels of auto-PEEP, which can lead to patient-ventilator dyssynchrony and increased work of breathing (figure 5 and figure 6) barotrauma, cardiovascular collapse, and potentially even death. (See "Positive end-expiratory pressure (PEEP)", section on 'Potential sequelae' and "Positive end-expiratory pressure (PEEP)", section on 'Auto (intrinsic) PEEP'.)

Auto-PEEP is common in patients with COPD. In a prospective cohort study of 13 patients with COPD who were being mechanically ventilated, all of the patients had measurable auto-PEEP (mean 9.4 cm H₂O), and seven had an auto-PEEP greater than 10 cm H₂O [42]. Auto-PEEP is responsible for up to one-third of the total work of breathing in patients mechanically ventilated with COPD [43].

Auto-PEEP can be detected in a number of ways. One practical and reliable method in patients with COPD is the demonstration on ventilator graphics of a progressive rise in peak airway pressures during mandatory volume-targeted ventilation [44-47]. Alternatively, ventilator time-flow graphics may demonstrate the commencement of inspiratory flow before expiratory flow reaches zero (figure 7 and figure 4). These methods are not quantitative. While auto-PEEP can be quantitatively assessed by measuring airway opening pressure during an end-expiratory pause (Paw) (waveform 2) [48], this method is only accurate when the patient is paralyzed or exhibiting negligible abdominal and chest wall muscle engagement during exhalation, which is uncommon in COPD patients requiring mechanical ventilation [44]. (See "Positive end-expiratory pressure (PEEP)", section on 'Assessment'.)

Prevention and treatment — In patients with COPD who are mechanically ventilated, the prevention and treatment of DHI are primarily achieved through a reduction in respiratory rate and/or tidal volume, raising the inspiratory flow rate (both prolonging expiratory time), and treating the underlying airflow obstruction when present [16]. In addition, the application of extrinsic PEEP can potentially reduce the work of breathing (WOB) (figure 8) and improve patient-ventilator synchrony through a reduction in ineffective triggering attempts by the patient (figure 6) [42,46,49]. Thus, the simultaneous manipulation of settings to achieve this balance should reduce ventilatory dyssynchrony and work of breathing, as well as decrease auto-PEEP and its attendant complications [19,25,34,47]. (See "Positive end-expiratory pressure (PEEP)", section on 'Treatment' and 'Ancillary therapy' below.)

When addressing DHI in a mechanically ventilated patient with COPD, the following should be considered:

●Reduce minute ventilation and prolong expiratory time – Because the main determinants of auto-PEEP are minute ventilation (Ve) and expiratory time (Te), the key strategies to limiting auto-PEEP focus on reducing total Ve, prolonging expiratory time, and improving expiratory flow [45,50-53]. This is primarily achieved through a reduction in respiratory rate and/or or tidal volume (the determinants of Ve) and raising the inspiratory flow rate (shortens inspiratory time [Ti] and prolongs Te) (figure 2). The consequent practice of permissive hypercapnia in this approach is acceptable, provided the arterial pH does not fall below 7.20 or lead to hemodynamic instability. (See 'Tidal volume' above and 'Ventilator rate' above and 'Inspiratory flow rate' above.)

●Adjust inspiratory trigger sensitivity to reduce ineffective triggering and lower the degree of wasted work of breathing (WOB) – Ineffective triggering and consequently high WOB and dyssynchrony are common in states of severe airflow limitation such as COPD. Mitigating patient-ventilator dyssynchrony and high WOB relies primarily upon optimizing the trigger sensitivity for inspiration [19,20,22,54]. In patients with COPD, ineffective triggering can be reduced by making sure the patient's intrinsic inspiratory time closely matches the duration of the tidal breath delivered by the ventilator in order to avoid active exhalation against the machine-driven breath (waveform 1) [20,25]. Inspiratory work can be reduced by adjusting the trigger sensitivity (usually flow triggered) to the lowest level needed to initiate a breath yet is ideally high enough to avoid false-triggering and overventilation (also known as "auto-cycling") (figure 3) [47,55-57]. (See 'Trigger sensitivity' above.)

●Apply extrinsic PEEP – In patients mechanically ventilated with COPD, the application of extrinsic PEEP has been shown to decrease WOB and improve patient-ventilator synchrony (figure 5 and figure 8) although it may not decrease DHI per se [46,58,59]. This is primarily due to a decrease in the energy required by the patient to reduce the airway pressure to trigger ventilator-assisted breaths. As an example, consider a patient whose trigger value is set to -2 cm H₂O and whose auto-PEEP is 10 cm H₂O. Without extrinsic PEEP, a breath will be triggered at an airway pressure of -2 cm H₂O, which requires that the patient decrease the alveolar pressure to the same level (by 12 cm H₂O). In contrast, when extrinsic PEEP of 5 cm H₂O is applied, a breath will be triggered at an airway pressure of 3 cm H₂O (ie, a -2 cm H₂O pressure differential below 5 cm H₂O), which requires that the patient decrease the alveolar pressure by only 7 cm H₂O. (See 'Applied PEEP' above and "Positive end-expiratory pressure (PEEP)", section on 'Applied (extrinsic) PEEP'.)

Although DHI can be prevented and treated with the above maneuvers, occasionally, patients present with life-threatening auto-PEEP causing cardiovascular collapse from reduced venous return. Severe auto-PEEP should be suspected in any intubated COPD patient who has acute hypotension in the setting of a rapid breathing rate or rising peak airway pressures. Importantly, when suspected, immediate disconnection of the endotracheal tube from the ventilator is a potentially life-saving maneuver, and is the most effective means of testing for and abolishing hemodynamically significant auto-PEEP.

ANCILLARY THERAPY — In addition to optimized ventilator settings, patients must be treated for the primary cause of their respiratory failure. Patients with severe obstruction from acute exacerbations of chronic obstructive pulmonary disease (AECOPD) should be aggressively treated with bronchodilators, glucocorticoids and antibiotics. Acetazolamide for the elimination of carbon dioxide does not appear to be of benefit [60]. (See "COPD exacerbations: Management" and "Treatment of community-acquired pneumonia in adults in the outpatient setting" and "Treatment of hospital-acquired and ventilator-associated pneumonia in adults".)

Bronchodilators — Patients with COPD who require mechanical ventilation should be given aerosolized bronchodilators to relieve acute and/or chronic obstruction. Beta-agonist and anti-cholinergic bronchodilators can be effectively delivered via metered dose inhalers (MDIs) or nebulizers during invasive mechanical ventilation using devices that can be attached to the inspiratory limb of the ventilator [61-65]. When compared to nebulized bronchodilators, MDI delivery results in sufficient bronchodilation when administered in higher than usual doses (eg, 4 to 8 puffs). MDIs are typically preferred because they are easy to use and are potentially less likely to result in ventilator tube contamination and ventilator-associated pneumonia.

The principles of bronchodilator use in patients mechanically ventilated with COPD are similar to that in patients with stable COPD or acute AECOPD, which are discussed separately. (See "Delivery of inhaled medication in adults", section on 'Mechanically ventilated patients' and "COPD exacerbations: Management", section on 'Emergency department and hospital management'.)

Glucocorticoids — Glucocorticoids are not routinely administered to patients who are mechanically ventilated with COPD unless they have an AECOPD. In AECOPD, systemic glucocorticoids reduce inflammation and relieve obstruction. While glucocorticoids have been shown to benefit nonventilated patients, the same mortality benefit has not been clearly reported in mechanically ventilated patients. A multicenter randomized study of 354 patients who required ventilatory support for AECOPD (noninvasive and invasive mechanical ventilation) demonstrated that compared to placebo, the administration of systemic glucocorticoids resulted in lower rates of NIV failure (0 versus 37 percent), shorter duration of mechanical ventilation (3 versus 4 days) and reduced length of ICU stay (6 versus 7 days) [66]. A single center trial demonstrated reduced additional days to weaning from mechanical ventilation (4.6 versus 7.2 days) and improved respiratory compliance in "difficult to wean" COPD patients (enrolled on average at day 6 of mechanical ventilation) using scheduled nebulized budesonide [67].

The dose and route of administration of glucocorticoids are similar in mechanically ventilated and non-mechanically ventilated COPD patients. The management of AECOPD with glucocorticoids is discussed separately. (See "COPD exacerbations: Management", section on 'Glucocorticoids in moderate to severe exacerbations'.)

Antibiotics — Antibiotics are not routinely administered for patients who are mechanically ventilated for COPD unless they have an acute exacerbation of COPD (AECOPD) or another indication for antibiotics, such as pneumonia. The rationale for this strategy for patients with AECOPD is derived from randomized studies and meta-analyses that report mortality benefits associated with antibiotic use, as well as reduced rates of readmission and shorter duration of mechanical ventilation. However, similar studies also report increased rates of infection with more virulent bacterial pathogens, as well as increased readmission rates for Clostridium difficile infection.  

The use of antibiotics in AECOPD is discussed separately. (See "Management of infection in exacerbations of chronic obstructive pulmonary disease".)  

Acetazolamide — Metabolic alkalosis, both acute and chronic compensatory, is commonly encountered among patients with AECOPD on invasive mechanical ventilation (IMV), either due to chronic carbon dioxide (CO₂) retention, or interventions commonly encountered in the ICU, such as nasogastric suctioning, diuretic and/or glucocorticoid use, or even permissive hypercapnia. Concerns for metabolic alkalosis prolonging the weaning process via central respiratory drive suppression has led many clinicians to employ acetazolamide in an effort to reverse metabolic alkalosis (via renal bicarbonate loss) and stimulate respiration. However, supporting data have been lacking for this practice. A case control study demonstrated no reduction in mechanical ventilation days using 500 mg/day of acetazolamide [68]. After pharmacodynamic modeling suggested that higher doses might be necessary [69], a multi-center, randomized clinical trial was undertaken to explore the use of higher dose acetazolamide, but 500 to 1000 mg twice daily again failed to demonstrate any significant reduction in duration of IMV [60].

High flow nasal cannula oxygen — There has been a growing trend to utilize high flow nasal cannula oxygen as an ancillary treatment for avoiding reintubation following liberation from mechanical ventilation. Proposed mechanisms of benefit include wash-out of the anatomical deadspace and mitigation of carbon dioxide rebreathing, increased nasopharyngeal pressure and intrinsic PEEP, improved conductance and compliance, and attenuated inspiratory resistance [70-72]. Demonstrated benefits in patients with COPD have included reduced respiratory rates, increased tidal volumes, reduced minute ventilation, reduced pCO₂ and transcutaneous CO₂, and a reduction in the rapid shallow breathing index and work of breathing [73-75]. Compared with conventional low flow oxygen, reduced reintubation rates have been reported with the routine application of HFNC following extubation in patients at low risk for reintubation (excluding moderate to severe COPD) [76]. Compared with NIV, HFNC has been shown to be equivalent in reducing the need for reintubation within 72 hours among high-risk patients (20 percent of which had moderate to severe COPD) [77]. In a prospective cohort comparing HFNC to noninvasive ventilation (NIV) in COPD patients with moderate hypercapnic acute respiratory failure, the extubation failure rate with HFNC (28.2 percent) was found to be not statistically less than that of NIV (39.5 percent), but skin breakdown issues and required nursing airway interventions were significantly less common in the HFNC group [78]. The role of HFNC postextubation and technical aspects of high flow oxygen are discussed separately. (See "Heated and humidified high-flow nasal oxygen in adults: Practical considerations and potential applications" and "Extubation management in the adult intensive care unit", section on 'Low-flow versus high-flow oxygen' and "Extubation management in the adult intensive care unit", section on 'High-flow oxygen via nasal cannulae'.)

PROGNOSIS — Mechanical ventilation for acute respiratory failure in patients with chronic obstructive pulmonary disease (COPD) is associated with high rates of intensive care unit (ICU) mortality (37 to 64 percent) [79-83]. A one year retrospective study of 4791 patients demonstrated that among COPD patients with long-term oxygen dependence who received IMV for AECOPD, in-hospital mortality was 23 percent. Among those surviving to discharge, 27 percent were in a skilled nursing facility within 30 days, and within 12 months, 67 percent were readmitted at least once, and 59 percent were dead [84]. A smaller observational study of less severe COPD patients requiring IMV reported a slightly lower in-hospital mortality (17 percent), and a lower rate of readmission (21 percent [85]). In addition, the ability of patients with COPD to successfully be liberated from mechanical ventilation is poorly studied [86]. Prognosis appears to be affected by both the severity of the underlying COPD as well as the reason for mechanical ventilation. Poor prognostic factors in this population include failure of noninvasive ventilation, the presence of multiorgan failure and virulent pathogens such as Pseudomonas and Aspergillus spp. While earlier studies suggested that mortality was determined by the severity of the underlying disease (eg, multiorgan failure, acute respiratory distress syndrome, and high Acute Physiologic and Chronic Health Evaluation [APACHE] score), several studies have since reported that a diagnosis of COPD alone is also a risk factor for death in mechanically ventilated patients:

●A retrospective study of 670 patients reported a significantly higher mortality in COPD patients with a higher acute physiologic and chronic health evaluation (APACHE) score or active malignancy [85].

●In one retrospective study of 428 patients with acute respiratory failure due to community acquired pneumonia, compared to patients without COPD, patients with COPD had a higher rate of requiring mechanical ventilation (odds ratio [OR] 2.78, 95% CI 1.6-4.7) and of dying (OR 1.58, 95% CI 1.0-1.4) [80]. The highest mortality was observed in intubated patients with COPD who had failed noninvasive ventilation (50 percent).

●In one prospective observational study of 235 patients mechanically ventilated for respiratory failure, compared to patients without a diagnosis of COPD, a diagnosis of COPD was an independent predictor of mortality (hazard ratio, 2.1; 95% CI 1.1-3.9) [81].

●In two observational studies of patients with ventilator-associated pneumonia, compared to patients without COPD, patients with COPD had higher rates of ICU mortality (64 versus 28 percent), and a longer duration of mechanical ventilation (24 versus 13 days) and ICU stay (26 versus 15 days) [82,87].

●A prospective analysis of 279 patients with ICU-acquired pneumonia demonstrated that 90 day mortality was higher in patients with COPD (57 percent) compared to those without COPD (37 percent) [88].

Infection with emerging pathogens, such as Pseudomonas and Aspergillus spp, may also be associated with high mortality in patients mechanically ventilated for acute respiratory failure and COPD [80,89-93]. COPD has been associated with an increased risk of MRSA and pseudomonas pneumonia [94] and a higher rate of infection with Aspergillus species [88] among patients with healthcare-associated pneumonia. The emergence of these organisms in COPD may be in part due to the high rate of initial empiric antibiotic and glucocorticoid therapy that is common in this population [89,95,96]. In particular, when the infection with Aspergillus spp is detected and treated late in its course, the mortality is thought to range from 80 to 100 percent [89,97]. (See "Diagnosis of invasive aspergillosis" and "Treatment and prevention of invasive aspergillosis" and "Management of infection in exacerbations of chronic obstructive pulmonary disease".)

Prognosis of AECOPD and acute respiratory distress syndrome (ARDS) are discussed separately. (See "Acute respiratory distress syndrome: Prognosis and outcomes in adults" and "COPD exacerbations: Prognosis, discharge planning, and prevention", section on 'Prognosis after an exacerbation' and "COPD exacerbations: Management".)

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: Chronic obstructive pulmonary disease".)

SUMMARY AND RECOMMENDATIONS

●The decision to institute invasive mechanical ventilation in patients with acute respiratory failure complicating chronic obstructive pulmonary disease (COPD) is based upon a constellation of clinical signs and symptoms, all within the context of the patient's preferences for life support. Common indications for intubation in this population include respiratory distress, failure to respond clinically to noninvasive ventilation, and gas exchange impairment that is unresponsive to medical therapy and/or noninvasive ventilation (failure to oxygenate or ventilate). (See 'Indications' above.)

●The primary objectives of mechanical ventilation in patients with COPD include supporting oxygenation and ventilation, minimizing the work of breathing, and avoiding dynamic hyperinflation (DHI). (See 'Goals of ventilator management' above.)

●For patients who require mechanical ventilation for acute respiratory failure complicating COPD, we suggest volume-targeted modes of ventilation over pressure support or pressure-limited modes of ventilation. Among the volume-targeted modes, we prefer volume-controlled modes (eg, volume control-continuous mandatory ventilation) or synchronized intermittent mandatory ventilation with pressure support [SIMV/PSV] if volume-controlled modes result in hyperventilation. Switching modes is not uncommon during the course of a patient's illness, particularly when complications of mechanical ventilation occur. (See 'Ventilator modes' above.)

●The ideal ventilator settings in patients with acute respiratory failure complicating COPD are those that meet the goals of mechanical ventilation. The initial ventilator settings will vary according to the cause of acute respiratory failure as well as the mode of mechanical ventilation selected. Subsequent adjustments should be made based upon the patient's clinical status, comfort level, gas exchange, and propensity for developing dynamic hyperinflation (DHI). Suggested initial ventilator settings are provided above. (See 'Initial ventilator settings and monitoring' above.)

●Intrinsic PEEP (also known as auto-PEEP) is common in patients with COPD who are mechanically ventilated, due to a high prevalence of spontaneous or ventilator-induced dynamic hyperinflation (DHI). Auto-PEEP can result in patient-ventilator dyssynchrony, increased work of breathing, barotrauma, cardiovascular collapse, and even death. (See 'Dynamic hyperinflation' above.)

●The prevention and treatment of DHI is primarily achieved through a reduction in respiratory rate and/or or tidal volume, raising the inspiratory flow rate, the application of extrinsic PEEP, and treating underlying airflow obstruction (if present) with bronchodilators and glucocorticoids. (See 'Prevention and treatment' above.)

●Patients with COPD who are mechanically ventilated should also be treated for the primary cause of their respiratory failure. This is especially important for patients with severe airflow obstruction from acute exacerbations of COPD, who should be aggressively treated with bronchodilators, glucocorticoids and antibiotics. (See 'Ancillary therapy' above.)

●The mortality of patients with COPD who are mechanically ventilated for acute respiratory failure is high (37 to 64 percent). Factors that portend a poor prognosis in this population include failure to respond to noninvasive ventilation, the presence of multiorgan failure, active malignancy, a high acute physiologic and chronic health evaluation (APACHE) score, and the presence of virulent pathogens such as Pseudomonas and Aspergillus species cultured from airway secretions. (See 'Prognosis' above.)