Authors: Jennifer C. Szafran, Bhakti K. Patel
Categories: Article, Mechanical ventilation, Assist control, Respiratory mechanics, Plateau pressure, Tidal volume, Positive end-expiratory pressure (PEEP)
Source: Critical care clinics
Authors: Jennifer C. Szafran, Bhakti K. Patel
Invasive mechanical ventilation may be indicated in cases of severe gas exchange abnormalities or an absence of airway protection. The precise indications vary, and the method of mechanical ventilation should be adapted to the cause of respiratory failure and the individual patient’s physiology. The goals of mechanical ventilation are to maintain sufficient gas exchange and to minimize iatrogenic harm while allowing time for lung recovery.
In cases of acute respiratory failure, intubations are often emergent with minimal time for preparation. In this context, many factors can influence the intubation strategy, with rapid sequence intubation techniques often used to maximize speed of intubation and minimize aspiration risk.^1,2^ In certain cases, including anticipated difficult airways, high aspiration risks, and severe physiologic derangements that may prevent the patient from tolerating apnea or induction medications, awake intubations can be considered. Regardless of the technique, consideration should be given to a backup plan before an initial attempt. The 2022 American Society of Anesthesiologists Practice Guidelines for Management of the Difficult Airway provide an algorithm to aid in identifying patients in whom to perform an awake intubation and in troubleshooting difficult airways.^3^
Recent emerging data surrounding intubation have focused on strategies for oxygen delivery and hemodynamic support. In critically ill adults, bag valve mask ventilation after induction and before laryngoscopy, as compared with no bag valve mask ventilation, has been associated with a lower incidence of periintubation severe hypoxemia without a significant increase in operator-reported or imaging-identified aspiration.^4^ Two trials have demonstrated no effect of a routine preintubation fluid bolus in critically ill adults on the incidence of periintubation cardiovascular collapse.^5,6^ (Table 1 for summaries of key randomized control trials pertinent to mechanical ventilation.)
Once the patient is intubated, the initial medication strategy should focus on analgesia to treat any pain with sedation as a supplement to treat any residual agitation. Of note, if the paralytic used during intubation is not reversed, the patient must be deeply sedated for the duration of the paralytic. The 2018 Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep Disruption Guidelines include key principles that can guide the initial sedation strategy. For example, for critically ill adults, propofol or dexmedetomidine is recommended over benzodiazepines (conditional, low-quality evidence).^7^
Modes of mechanical ventilation can be distinguished based on breath initiation and breath delivery. (Table 2 highlights key differences in ventilator modes.) Breaths can be initiated by the ventilator, the patient, or a combination of the two. A key dichotomy in many modes of ventilation is whether breaths are delivered with a set amount of volume or a set amount of pressure. In either case, the other (dependent) variable reflects the compliance of the respiratory system (Crs) and can thus provide critical information about pathophysiology.
In other words, if volume is set on the ventilator, pressure is the dependent variable monitored to understand compliance. If pressure is set on the ventilator, volume is the dependent variable monitored to understand compliance.
Assist control (AC) methods are common forms of invasive mechanical ventilation in the medical intensive care unit (ICU). In these modes, the ventilator guarantees a minimum number of breaths (the set respiratory rate) and a set amount of volume (assist control volume [AC/V]) or pressure (assist control pressure [AC/P]) delivered for each breath. If the patient breathes over this set rate, the breaths delivered will be identical to those breaths initiated by the ventilator. This is in contrast to a control mode of ventilation, in which the ventilator will not deliver any breaths triggered by the patient. (However, in clinical practice, few make this distinction as AC is typically used.)
Additional variables set by the provider include fraction of inspired oxygen (Fio2), which is set in all modes of ventilation, and positive end-expiratory pressure (PEEP). In AC/V, both a peak flow and a flow pattern can be set. These, in combination with the set tidal volume, determine the inspiratory and expiratory time. In AC/P, an inspiratory time or an expiratory ratio can be set.
As described in the “Respiratory Mechanics” section, AC/V is particularly useful for obtaining and interpreting respiratory mechanics. It is well studied in acute respiratory distress syndrome (ARDS) with resultant evidence-based parameters. Additionally, because both tidal volume and minimum respiratory rate are set by the provider, a minimum minute ventilation can be guaranteed. However, particularly with low tidal volume strategies, ventilator dyssychrony can result. AC/P may result in a more natural flow pattern and thus improved work of breathing^8^ and ventilator synchrony in some cases. However, in general, compared with AC/V, a longer inspiratory time is needed to deliver the same volume of breath in AC/P. This increased inspiratory time results in an increased risk of autoPEEP. Additionally, because tidal volume is a dependent variable not set by the operator in AC/P, minute ventilation cannot be guaranteed. Therefore, close monitoring is required to ensure consistent ventilation.
Support modes of ventilation rely on respiratory effort from the patient. In pressure support ventilation (PSV), the provider sets the PEEP and the pressure support (Ps), which is the pressure provided in addition to the PEEP during inspiration. An increased Ps without a change in PEEP will result in an increased driving pressure and thus, typically, an increased tidal volume. In AC/P, during inspiration, patients receive the chosen pressure for a set amount of time. In contrast, in PSV, patients receive the set amount of pressure until there is a prespecified change in their inspiratory effort (based on a change in flow). Although there is a backup mode of ventilation set in case of emergencies, in PSV, all breaths must be initiated by the patient. This makes PSV a good choice of mode in a patient who has normal lung function and a normal respiratory drive but is intubated for airway protection, allowing them diaphragmatic exercise and potentially less analgesia/sedation. In fact, some sedation strategies are not compatible with PSV if they significantly decrease the respiratory drive. However, the effort/energy expenditure required for the patient to control all the breaths in PSV may not be desired depending on the nature of the patient’s critical illness.
Synchronized intermittent mandatory ventilation (SIMV) combines AC modes of ventilation with PSV. The provider sets a certain number of breaths to be delivered by volume control (intermittent modes using pressure control also exist). Any additional patient-initiated breaths are delivered as pressure-supported breaths. Because of these pressure-supported breaths, SIMV was originally assumed to have exercise benefits for weaning. However, it may actually prolong time to extubation when compared with some other methods, including intermittent spontaneous breathing trials and PSV, depending on the criteria used.^9,10^
Pressure regulated volume control (PRVC) is a pressure control mode of ventilation that automatically adjusts to meet a target tidal volume. The provider sets a rate, target tidal volume, PEEP, and maximum pressure control. During inspiration, the machine delivers a level of pressure and then adjusts future breaths based on the observed tidal volume. For example, if the observed tidal volume is less than the set target tidal volume, during the next inhalation, the ventilator will deliver more pressure. If the observed tidal volume is more than the set target tidal volume, during the next inhalation, the ventilator will deliver less pressure. The benefit of this mode is that it can adapt to a patient’s evolving physiology. However, this is also a significant potential downside. Because the machine adjusts automatically, a provider may not be alerted/notice if an important physiologic change has taken place (eg, if more pressure is needed to maintain the same tidal volume). Additionally, the machine may have trouble successfully adjusting if there is significant ventilator dyssynchrony. Finally, as with AC/P, there is often increased risk of autoPEEP due to longer time to deliver breaths when compared with AC/V.
Airway pressure release ventilation (APRV) is sometimes considered in patients that are very difficult to oxygenate. The goal of APRV is to maximize lung recruitment while minimizing potential ventilator-associated lung injury due to repeated opening and closing of alveolar units. To achieve this, the mode allows for spontaneous breathing at a high PEEP (Phigh) for time Thigh alternating with a relatively short period of exhalation (Tlow) at a low PEEP (Plow). Risks include those associated with long periods of elevated intrathoracic pressures (including decreased venous return) and a risk of autoPEEP. There is much heterogeneity in the ARDS and trauma literature regarding APRV, making it difficult to form a definitive conclusion regarding its efficacy compared with other modes of ventilation^11–15,16–22^ with some demonstration of increased ventilator-free days in adults with ARDS^14^ but one study demonstrating a trend toward increased mortality in a pediatric ARDS population.^17^
High-frequency modes of ventilation are rarely used in adults as an attempted rescue therapy for patients who are unable to achieve adequate gas exchange with conventional modes of ventilation. As the name implies, this mode delivers breaths at very high rates (multiples above physiologic respiratory rates) and very low tidal volumes with a set mean airway pressure. Due to the low tidal volumes, one potential benefit is the minimization of volutrauma. The most studied mode in this category is high-frequency oscillatory ventilation (HFOV), in which the provider selects a frequency (rate), amplitude, mean airway pressure, flow, I:E ratio, and Fio2. However, HFOV has been demonstrated to result in no difference^18^ or an increase^19^ in mortality as compared with more standard methods of ventilation in patients with moderate or severe ARDS and should thus be used with caution.
Although there are many modes of mechanical ventilation available, using diagnostics within one consistent mode allows for minute-to-minute understanding of a patient’s evolving pathophysiology. Respiratory mechanics are measurements of pressure, volume, and/or flow that provide information about the function of the respiratory system. They are traditionally measured in AC/V mode with a square waveform (flow is either zero or a constant rate, as opposed to a decelerating waveform).
In AC/V, the volume and flow are selected by the provider, and thus, the pressure is the dependent variable that is measured. The peak airway pressure (Ppeak) is the highest pressure in a respiratory cycle, reached at the end of inspiration. It is the sum of resistive pressure (Pres), elastic pressure (Pel), and PEEP.^20^
Ppeak=Pres+Pel+PEEP
Pres is proportional to flow and resistance of the respiratory system (Rrs) and is thus elevated, for example, in obstructive lung diseases.
Pel is inversely proportional to Crs and is thus elevated, for example, in restrictive lung diseases and obesity. Of note, the Crs includes the compliance of both the lung and the chest wall.
Pel can be calculated by measuring a plateau pressure (Pplat). Pplat is the sum of Pel and PEEP. It is determined at the bedside via an inspiratory hold (Fig. 1). At the end of inspiration, the provider stops the flow of air. At this time, the pressure, with some limitations, should be representative of the pressure at the alveoli.
If the provider has measured Ppeak, Pplat, and PEEP (set by provider), the Pres and Pel can be calculated. A normal value for Pres is about 10 cm H2O or less. If abnormal, these values can be helpful in determining whether the patient has obstructive or restrictive pathophysiology on initial mechanical ventilation. Importantly, serial measurements at stable tidal volumes and flow can help to diagnose new problems that develop (Fig. 2).
Respiratory mechanics were measured for Patient A during mechanical ventilation on AC/V with a square flow waveform. The following values were Ppeak of 25 cm H2O, Pplat of 15 cm H2O, and a PEEP of 5 cm H2O (see Fig. 2A). Later in the day, the provider is called to the patient’s bedside because the Ppeak suddenly increased to 45 cm H2O. There have been no changes to the settings on the ventilator. The provider remeasures the pressures.
If the Ppeak has increased but the Pplat has not, then there is a new issue with airways resistance or flow. In this case, the numbers measured for Patient A are Ppeak = 45 cm H2O, Pplat = 15 cm H2O, and PEEP = 5 cm H2O. The calculated Pres has dramatically increased (from 10 cm H2O to 30 cm H2O), suggesting an issue with airways resistance or flow (see Fig. 2B). When developing a differential, a good strategy is to think through (and assess at bedside) the circuit starting at the mouth to identify potential sources of resistance to flow. Examples include patient biting on tubing or kinking of the tubing, mucus in tubing, and bronchospasm.
If the peak airway pressure has increased, and the Pplat has increased similarly, there is a new issue with respiratory system compliance. In this case, the numbers measured for Patient A are Ppeak = 45 cm H2O, Pplat = 35 cm H2O, and PEEP = 5 cm H2O. The calculated Pel has dramatically increased (from 10 cm H2O to 30 cm H2O), suggesting an issue with respiratory system compliance (see Fig. 2C). Examples that happen in the acute setting include the movement of the endotracheal tube into one lung (and thus new single lung ventilation), pneumothorax, flash pulmonary edema, large mucus plugging causing significant atelectasis, and compartment syndrome in the abdomen or chest.
Another important measurement on the ventilator is the end expiratory pressure (Fig. 3). This is measured by performing an end expiratory hold at the end of exhalation (holding flow at zero). This measured pressure should match the PEEP that the provider set on the ventilator. If the pressure measured during and end expiratory hold is higher than the PEEP, this indicates a level of intrinsic PEEP. This can result from the autoPEEP due to a patient repeatedly being unable to fully exhale before the next breath (eg, in the case of status asthmaticus). Significant autoPEEP is important to identify because it can result in decreased gas exchange, shock, and circulatory arrest from decreased venous return.
Although many general principles can be applied to all patients with respiratory failure undergoing mechanical ventilation, there are some conditions that require a tailored approach.
Mechanical ventilation strategies in ARDS have been frequently studied. As a result, ARDS is the only type of acute respiratory failure with multiple guidelines^21,22^ outlining evidence-based strategies for mechanical ventilation. Patients with ARDS are at particularly high risk of volutrauma and barotrauma, consistent with the characteristic diffuse alveolar damage and subsequent worsening lung compliance. In this context, there is significant potential for iatrogenesis during mechanical ventilation. As a result, a lung-protective ventilation strategy that includes low tidal volumes (4–8 mL/kg predicted body weight [PBW]) and a target Pplat of 30 cm H2O or less is used.
Exacerbations of asthma and chronic obstructive pulmonary disease (COPD) can be particularly difficult to manage on the ventilator. Patients often require a significantly increased expiratory time to fully exhale. Failure to fully exhale can result in autoPEEP. This can result in decreased gas exchange, shock, and even circulatory arrest from decreased venous return. To avoid this, inspiratory time should be minimized so that expiratory time can be maximized. The most effective strategy to accomplish this is reducing the respiratory rate. Due to patient tachypnea, achieving this reduction in respiratory rate sometimes requires deep sedation and even paralysis to achieve. In AC/V, increasing peak flow rate and selecting a square flow waveform will minimize inspiratory time (and predictably increase peak airway pressures). In AC/P, an inspiratory time can be selected by the physician. Acutely, if autoPEEP results in hemodynamic compromise, the patient should be disconnected from the ventilator, brief pressure should be applied to the chest to minimize any remaining volume, and the patient should be reconnected to the ventilator with new settings that decrease the expiratory ratio. If the hemodynamic compromise was indeed from auto-PEEP, the improvements should be instantaneous but this maneuver may need to be repeated if the hemodynamic changes recur.
Before adjusting variables on the ventilator, it is important to understand our goals. These goals can include target values for dependent variables, as well as lung protection.
Minute ventilation is a product of tidal volume and respiratory rate. Therefore, these 2 variables influence partial pressure of CO2 (Pco2) and thus pH. In contrast, PEEP and Fio2 primarily influence partial pressure of oxygen (Po2) and oxygen saturation (SpO2). Therefore, targets for pH and Po2 help guide titration of these variables on the ventilator.
In part because patients may have varying levels of baseline Pco2, a pH target (rather than a Pco2 target) is generally used to guide ventilation. A relatively normal pH is a reasonable target when possible in the context of lung protective ventilation. However, permissive hypercapnia and the resulting respiratory acidosis are generally well tolerated if needed.^23^ This is often required in cases of ARDS or severe exacerbations of reactive airways disease. Though there is not a consensus as to the lower limit of tolerated pH, pH ≥ 7.2 is a reasonable goal. It is similarly important to avoid severe alkalosis, which can have consequences that include arrhythmias and seizures. Alkalosis can also decrease or eliminate respiratory drive, which can decrease success with spontaneous breathing trials.
Oxygenation targets remain a topic of controversy. Although initial single-center data of medical-surgical ICU patients suggested that a conservative oxygen strategy (partial pressure of oxygen in arterial blood [Pao2] between 70 mm Hg and 100 mm Hg or SpO2 between 94% and 98%) resulted in lower ICU mortality as compared with controls,^24^ subsequent trials (using different O2 targets) in mechanically ventilated patients,^25^ patients with ARDS,^26^ and ICU patients with acute hypoxic respiratory failure^27^ have not demonstrated a consistent difference in outcomes. This remains an area of active investigation.
A low tidal volume strategy (4–8 mL/kg PBW) has been demonstrated to decrease mortality and increase ventilator free days when compared with a high tidal volume strategy (initially 12 mL/kg PBW) in ARDS^28^ and is recommended by multiple ARDS guidelines.^16,17^ This strategy is often extrapolated to the management of mechanically ventilated patients with other pathology. This is supported by some data. For example, data from surgical patients demonstrated that, even in the intraoperative period, lung protective ventilation resulted in a decreased incidence of clinical complications.^29^ A benefit of a low tidal volume strategy versus an intermediate tidal volume strategy has not been demonstrated. In the PReVENT trial of mechanically ventilated patients without ARDS, those receiving a low tidal volume strategy (initial volume of 6 mL/kg PBW) as compared with those receiving an intermediate tidal volume strategy (initial volume of 10 mL/kg PBW) did not demonstrate increased ventilator-free days.^30^ In patients with severe obstructive lung disease exacerbations that require very low respiratory rates or in patients with severe metabolic acidosis, a more liberal tidal volume strategy (8–10 mL/kg of PBW) may be required to maintain a safe pH.
Mechanically ventilated patients with respiratory failure can have a variety of factors that lead to an increased required minute ventilation (eg, increased metabolic demand and increased dead space) and thus an increased required respiratory rate. Additionally, a lung protective tidal volume strategy usually results in tidal volumes lower than that of the typical spontaneous breath in a healthy person. As a result, a high respiratory rate is often required. This is particularly true in ARDS (low tidal volume strategy, increased dead space) and patients with shock (the ventilator must often compensate for increased metabolic acidosis.) Whenever a set respiratory rate on the ventilator is increased, the provider should ensure no autoPEEP results. This is particularly true in patients with obstructive lung disease, who might require very low respiratory rates to prevent autoPEEP. When respiratory alkalosis is identified, the provider should ensure it is not iatrogenic (ie, the patient is breathing at the set rate and tidal volume, which have been set too high, resulting in too high of a minute ventilation). If the alkalosis is not iatrogenic (ie, the patient is breathing over the set respiratory rate), the cause of the tachypnea should be identified and addressed.
Goals of PEEP include recruiting alveoli and preventing ventilator-associated lung injury due to repeated alveolar collapse. Although an active area of investigation for many years, there is no consensus regarding the best strategy for choosing PEEP.^31–33^ However, generally, a lung-protective ventilation strategy includes a goal Pplat of less than 30 cm H2O. Additionally, evidence suggests strategies that include recruitment maneuvers that use very high levels of PEEP (25–45 cm H2O) result in increased mortality.^34^
Other strategies for PEEP adjustment currently under investigation include targeting lower driving pressures (calculated equivalently to Pel as Pplat – PEEP)^35–38^ with one reasonable target less than 15 cm H2O. Esophageal monitoring can be used to measure transpulmonary pressures and has been implemented with mixed results.^39,40^ The idea behind this strategy is to account for situations, such as decreased chest wall compliance due to obesity, when plateau pressures may not be an accurate representation of transpulmonary pressures. Other potential considerations requiring more investigation include pressure–volume curves, stress index, and electrical impedance tomography.^41^
Very high levels of Fio2 raise concern for toxicity related to reactive oxygen species. For this reason, the goal is using the lowest Fio2 needed to achieve a set SpO2 or Pao2 target, with Fio2 of 60% or less preferred when possible.^42^
Both peak inspiratory flow rate and flow waveforms can be set. These rarely need to be manipulated. However, flow does contribute to peak airway pressures, so it should remain constant between serial measurements of respiratory mechanics. Similarly, respiratory mechanics are traditionally calculated with a square flow waveform (flow is either zero or a constant value) on AC/V. One condition that may benefit from manipulation of flow is an obstructive lung disease exacerbation. These patients may benefit from an increase in flow (with monitoring of resulting airway pressures) to decrease the expiratory ratio and allow longer time to exhale.
The provider can set the ventilator to allow a patient to trigger a breath by generating either a certain flow or a certain pressure. This does not often have to be manipulated from standard settings. Patients with respiratory muscle weakness (especially when on a support mode of ventilation) may require adjustments to trigger settings to make it easier to trigger a breath. In contrast, if trigger threshold is set too low (based on absolute value), breaths could be triggered by nonrespiratory stimuli (cardiac pulsation, condensation in the tubing, myoclonus, and so forth).
Even when ventilator settings are carefully chosen based on the considerations listed above, if a patient maintains a respiratory drive, the “desired” breathing pattern may differ significantly from the ventilator settings. This can result in perturbations to the expected delivered breath referred to as ventilator dyssynchrony. Ventilator dyssynchrony puts the patient at additional risk of ventilator-associated lung injury (including volutrauma and barotrauma). Perhaps for this reason, in observational studies, high ventilator dyssynchrony has been associated with longer duration of mechanical ventilation^43,44^ and higher mortality.^45^ Increases in sedation are not always effective at reducing ventilator dyssynchrony and have been shown to be less effective than making adjustments to ventilator settings.^46^ Therefore, initial efforts should focus on ventilator adjustments. There are many types of ventilator dyssynchrony. Three important, intervenable examples are listed as follows.
Breath stacking, or double triggering, results from a patient triggering a second breath before exhalation on a first breath is completed. This can result in increased tidal volumes and pressures. Strategies to improve this can include changing ventilator modes, with PSV often highlighted.^47^ However, caution should be used when switching to a spontaneous mode of ventilation to ensure that patients are still taking lung-protective breaths. High respiratory drive, even in patients who are not mechanically ventilated, can contribute to an injurious breathing pattern with associated patient self-inflicted lung injury.^48^ Another strategy to reduce breath stacking is to increase inspiratory time. In AC/V, this can be done by adding an inspiratory pause, changing flow (decreased flow will prolong inspiratory time, although some dyssynchrony results from insufficient flow), or increasing tidal volume (which may also decrease the patient’s sensation of “air hunger”). Any increase in inspiratory time will result in a decrease in expiratory time, so after these changes, the provider must confirm that no auto-PEEP has developed.
When a patient creates a respiratory effort (identified as a change in pressure and flow) that does not trigger a full breath on the ventilator, this is referred to as ineffective triggering. This can occur when the trigger threshold is not set at a level to detect the patient efforts or in the case of intrinsic PEEP.^36^ Therefore, it can often be improved with adjustments to the trigger threshold or, in the case of intrinsic PEEP, adjustments to the ventilator to reduce auto-PEEP or increase extrinsic PEEP.
Occasionally, dyssynchrony is not the result of respiratory efforts from the patient. In autocycling, the ventilator may detect a breath trigger resulting from a change outside of the patient’s respiratory muscles, including moving condensation in the ventilator tubing, myoclonus, or cardiac oscillations.^49^ A high index of suspicion is needed to identify this issue. The erroneous trigger should be eliminated if possible (eg, tubing exchanged). Adjusting the trigger threshold can be used if needed.
Because of the potential harms associated with mechanical ventilation, as soon as a patient is intubated, regular consideration should be given to when the patient can appropriately be extubated. This decision is based on a variety of factors, as reviewed in another article of this collection. Regardless of the cause of respiratory failure, it is important to assess at least daily, and often more frequently, whether the patient is appropriate for liberation.
Invasive mechanical ventilation allows clinicians to support the gas exchange and work of breathing of patients with respiratory failure. Although mechanical ventilation provides a potential bridge to lung recovery, there is also potential for iatrogenesis. Understanding modes of ventilation and the variables that can be manipulated in each facilitates the goals of sufficient gas exchange and avoiding harm. Ventilators also provide crucial diagnostic information that should be regularly reassessed throughout the duration of mechanical ventilation.