In certain circumstances, positive pressure ventilation can cause or contribute to acute lung injury.³ Patients most at risk for ventilator-induced lung injury are those with severe systemic (e.g., sepsis) or pulmonary (e.g., pneumonia) disease who are ventilated with large tidal volumes, high airway pressures, inadequate levels of PEEP, and high inspired oxygen concentrations.⁴ The mechanisms of ventilator-induced lung injury are complex but involve an exacerbation of systemic inflammation⁵ and physical trauma to the alveoli due to overdistension and the repeated opening and closing of lung units.⁶ Crude mortality rate in critically ill patients ventilated with high tidal volumes (12 ml/kg) is 9% higher than the rate in those ventilated with low tidal volumes (6 ml/kg).⁷ This observation has led to the concept of lung-protective ventilation, which is described subsequently.

Cardiac Effects

The main direct effects of positive ventilation on the cardiovascular system are reduced systemic venous return,⁸ altered right ventricular afterload,⁹ and reduced left ventricular afterload.¹⁰ The impact of these effects is determined by the particular patient’s underlying cardiac function and intravascular volume status. Positive pressure ventilation performed properly eliminates or substantially reduces the work of breathing, which also has an indirect positive effect on cardiovascular function.

Other Effects

Positive pressure ventilation results in reduced secretion of atrial natriuretic peptide, which leads to sodium and water retention by the kidney.¹² Patients who undergo prolonged positive pressure ventilation typically retain significant amounts of sodium and water.

VENTILATOR MODES AND CONTROLS

Ventilators have become very complex and have a bewildering array of modes and controls,¹³ despite a lack of evidence of improved outcomes.¹⁴ Further confusion is created because manufacturers call similar ventilatory modes different names.

At its most simple, invasive ventilation involves mandatory machine-delivered breaths of a given volume (or pressure), in which controls for fractional inspired oxygen (F_Io₂), breath rate, breath volume (or pressure), and the ratio of inspiratory time to expiratory time (the I:E ratio) are set.¹⁵ For this type of mandatory ventilation to succeed, the patient’s respiratory system must be entirely passive. Much of the complexity of modern ventilators comes from the desire to allow patients to take spontaneous breaths that are assisted by and synchronized with the ventilator.

The following parameters apply to all modern ventilators: (1) inspiratory flow generation; (2) cycling; (3) triggering; (4) mode; (5) PEEP.

Inspiratory Flow Generation: Volume Control or Pressure Control

Volume control ventilation provides a constant inspiratory flow to achieve a predefined tidal volume (Fig. 29-2). Airway pressure varies with inspiratory time and changes in lung compliance and airway resistance. When setting the ventilator in volume control mode, a tidal volume of 6 to 8 ml/kg should be chosen.

Figure 29.2 Pressure-time and flow-time curves for volume control and pressure control ventilation. A, Two volume control breaths are illustrated. The I:E ratio is approximately 1:2 and there is a short inspiratory pause at the end of inspiration. Plateau pressure is a slightly lower than peak pressure. Inspiratory flow is constant, whereas flow decreases during expiration. B, Two pressure control breaths are illustrated. The I:E ratio is approximately 1:2. No end-inspiratory pause is set. Inspiratory flow is decelerating and has fallen to zero by end-inspiration. In both A and B, expiratory flow has fallen to zero before the end-expiration. For the sake of clarity, the inspiratory rise time (see Fig. 28-4) has been set to zero; that is, peak inspiratory flow occurs immediately at the beginning of inspiration.

Pressure control ventilation provides a constant inspiratory pressure, which generates a decelerating inspiratory flow pattern (see Fig. 29-2). Tidal volume varies with inspiratory time and changes in lung compliance and airway resistance. The decelerating flow pattern distributes ventilation more evenly throughout the lung; in patients with V/Q mismatch, this improves oxygenation. It also allows some compensation for air leaks and limits inspiratory pressures, which helps to prevent ventilator-induced lung injury and barotrauma. The inspiratory pressure is chosen to achieve the desired tidal volume, usually 6 to 8 ml/kg, which typically requires a pressure of 15 to 30 cm H₂O.¹⁶

Volume control is commonly used for routine postoperative ventilation. It has the advantage of providing constant minute ventilation but the disadvantage of providing potentially high airway pressures. Pressure control ventilation is used in patients with reduced lung compliance or impaired gas exchange and has the advantage of limiting airway pressures but the disadvantage of providing variable minute ventilation.

Most modern ventilators incorporate hybrid methods of flow generation that produce a constant tidal volume with a decelerating flow pattern. These hybrid forms provide the benefits of pressure control ventilation (decelerating flow, pressure limitation) while ensuring a fixed tidal volume.

Cycling

Cycling refers to the mechanism by which the ventilator changes between inspiration and expiration. Cycling may be time-based or flow-based.

Time

Time cycling is used for mandatory modes of ventilation. Inspiratory and expiratory times are defined by setting the respiratory rate and either the inspiratory time or the I:E ratio. A breath rate of 10 to 15 per minute and an inspiratory time of 1.2 to 1.5 seconds (or an I:E ratio of 1:2 or 1:3) are appropriate in most circumstances. If the breath rate is set very low (<10/min), the I:E ratio should be decreased to 1:4 or 1:5 or lower. Varying the inspiratory time affects either tidal volume (pressure control mode) or airway pressure (volume control mode).

In patients with increased airway resistance, a reduced respiratory rate and prolonged expiratory time is required to avoid dynamic lung hyperinflation (see subsequent material).¹⁷ For instance, a breath rate of 6 to 8 per minute with an inspiratory time of 1.3 seconds provides an expiratory time of 6.2 to 8.7 seconds and an I:E ratio of 1:4.7 to 1:6.7, respectively.

In patients with severe hypoxemia, a prolonged inspiratory time may enhance oxygenation by improving the ventilation of lung units with long time constants (see Chapter 1). In some circumstances, inverse ratio ventilation (e.g., an I:E ratio of 2:1 or 3:1) may be used (see Ventilatory Strategies in subsequent material).¹⁸

Flow

For some assisted modes of ventilation, such as pressure support, flow cycling is used to terminate inspiration and commence expiration. Termination of inspiration occurs when flow drops to a percentage of peak inspiratory flow (typically 0 to 25%). Flow cycling may improve patient-ventilator synchrony.¹⁹

Triggering

Triggering is the mechanism by which a patient’s spontaneous inspiratory effort is sensed and the ventilator is triggered to deliver inspiratory assistance. Thus, triggering is a method of cycling from expiration to inspiration that is not time dependent. Triggering is accomplished when the ventilator senses a change in flow or pressure within the ventilator circuit. Flow triggering is generally more sensitive than pressure triggering and results in improved patient-ventilator synchrony.¹⁵

The trigger sensitivity should be set at a low level so that triggering requires minimal patient effort, but not so low that the ventilator triggers itself (autotriggering); the usual values are about 1 l/min for flow triggering or 1 cm H₂O for pressure triggering. Autotriggering is usually manifested by a rapid and regular respiratory rate. Autotriggering may be caused by an endotracheal cuff leak, movement of water within the circuit, a prominent cardiac impulse, or patient movement.

Ventilation Modes

The modes of ventilation form a confusing area. Reference to the manufacturer’s manual usually clarifies the terminology used in discussing a particular ventilator. Only three commonly used modes are discussed here.

Intermittent Mandatory Ventilation

With intermittent mandatory ventilation, a mandatory number of breaths of a set volume or pressure are delivered each minute. It can be used with volume or pressure flow generation and is time cycled. Patients may breathe between the preset breaths but no attempt is made to synchronize mandatory breaths with spontaneous breaths. With synchronized intermittent mandatory ventilation (SIMV), the ventilator attempts to synchronize the patient’s respiratory effort with the mandatory machine breaths. SIMV can be used with constant, variable, or hybrid inspiratory flow generation and is usually combined with pressure support.

Assist/Control

Assist/control can be used with volume or pressure control and is time cycled. If the patient makes no spontaneous respiratory effort, machine breaths are delivered at the set rate (control mode). Triggering the ventilator causes a machine-set breath to be delivered (assist mode) and leads to a reduction in the number of controlled breaths delivered. As a patient begins to breathe, the number of control breaths declines until the patient is breathing spontaneously, with assisted breaths only.

When assist/control is used with volume control, all breaths, both controlled and assisted, have the same tidal volume. When used with pressure control, all breaths have the same pressure. Both controlled and assisted breaths also have the same inspiratory time, which is set. Once the patient is breathing at or above the set breath rate, the breath rate setting no longer functions.

Pressure Support

In pressure support mode, spontaneous breaths are augmented to a preset pressure. The patient must trigger the ventilator to initiate inspiratory support, and the ventilator cycles from inspiration to expiration on the basis of a set percentage drop in the peak inspiratory flow rate (usually a fall of 75% to 100%). Thus, breath rate, inspiratory time, and tidal volume are variable and patient controlled. Provided that patients are able to trigger the ventilator, pressure support can provide full ventilation. However, there is the potential for apnea and rapid shallow breathing. To prevent these problems, pressure support is often combined with SIMV. The level of pressure support is normally 5 to 10 cm H₂O above PEEP but may occasionally be set higher.

Positive End-Expiratory Pressure

PEEP is positive pressure within the alveoli at endexpiration. It is applied by placing a pressure-limiting valve on the expiratory limb of the ventilator circuit. PEEP reduces extravascular lung water, prevents the opening and closing of small airways, and helps to recruit collapsed alveoli. This improves oxygenation, increases lung compliance, and reduces the risk of developing ventilator-induced lung injury.^20,²¹ A low level of PEEP (5 cm H₂O) should be applied to all patients who are invasively ventilated. High levels of PEEP (15 to 20 cm H₂O) may be required in patients with acute respiratory distress syndrome (ARDS).²² Patients with increased airways resistance whose ventilators are set to an inadequate expiratory time may trap gas and generate auto-PEEP. By applying external PEEP at a level similar to that of the generated auto-PEEP, the inspiratory threshold load (see Fig. 27-4), and therefore the work of breathing, is reduced.

Ventilator Pressures and Graphics

Important information about the patient’s respiratory status may be obtained from the ventilator pressure and graphic displays.

Measured Pressures

Peak Pressure and Plateau Pressure.

Peak pressure is the maximum inspiratory pressure within the breathing circuit. It is affected by both respiratory compliance and airway resistance. Plateau pressure is the pressure within the breathing circuit following an end-inspiratory pause, which allows equalization of any pressure difference between the alveoli and the circuit. Thus, plateau pressure is indicative of alveolar pressure at end inspiration and is influenced by respiratory compliance alone. Plateau pressure may be measured (the patient must be heavily sedated, paralyzed, or both) using the inspiratory hold function on the ventilator or, in volume control mode, estimating it from the pressure-time curve (see Fig. 29-3) when an inspiratory pause or hold is used. Plateau pressure (PP) may be calculated from the respiratory system compliance (C) and the tidal volume (Vt), given:

Figure 29.3 Pressure-time and flow-time curves showing the effects of reduced respiratory system compliance and increased airways resistance. A, With decreased respiratory compliance (middle breath), the peak inspiratory pressure increases, as does the slope of the airway pressure curve. The plateau pressure is also increased. Tidal volume (the area under the flow-time curve) remains constant. With increased airway resistance, peak inspiratory pressure is increased, but plateau pressure is not increased provided that a sufficiently long inspiratory pause is used. B, With decreased respiratory compliance, the peak inspiratory pressure remains constant but the inspired tidal volume (the area under the flow-time curve) has decreased. With increased airways resistance, peak inspiratory pressure is unchanged, but peak inspiratory flow is reduced and inspiratory flow has not fallen to zero by end-inspiration. Tidal volume is decreased. Expiratory flow rate is decreased and takes longer to reach zero.

(29-1)

Ventilator-induced lung injury is minimized by limiting plateau pressure to less than 32 cm H₂O (and peak airway pressure to less than 35 cm H₂O).⁴ However, when chest wall compliance is reduced (e.g., in cases of obesity or abdominal compartment syndrome), a higher plateau pressure may be tolerated (see Fig. 29-1).

With volume control ventilation, high peak and plateau pressures may be due to reduced lung or chest wall compliance (see Figs. 29-1 and 29-3), whereas high peak but normal plateau pressure is indicative of increased airway resistance (see Figs. 29-1 and 29-3).

End-Expiratory Pause Pressure and Auto-PEEP.

End-expiratory pause pressure is the pressure within the circuit following an end-expiratory pause, and it should equal the level of applied PEEP. If end-expiratory pause pressure is greater than the applied PEEP, auto-PEEP exists (Fig. 29-4).²³ Auto-PEEP is caused by continued (delayed) emptying of alveoli at end-expiration and indicates increased airway resistance. The presence of auto-PEEP is suggested by continued gas flow at end-expiration on the flow-time curve (see Fig. 29-4).

Figure 29.4 Pressure-time and flow-time curves showing auto-PEEP due to increased airway resistance. A, With volume control ventilation, peak inspiratory pressure and the slope of the pressure-time curve increase from breath 1 to breath 4. This reflects the reduced compliance of the lungs as they become overdistended. Tidal volume remains constant and expiratory flow does not reach zero by the end-expiration. An expiratory hold following the fourth breath allows equilibration between the circuit and the patient and shows the level of auto-PEEP. B, With pressure control ventilation, peak inspiratory pressure remains constant but tidal volume (the area under the flow-time curve) progressively falls between breath 1 and breath 4. This reflects the reduced compliance of the lungs as they become overdistended. Despite this reduced tidal volume, end-expiratory flow does not reach zero. An end-expiratory hold following the fourth breath allows equilibration between the circuit and the patient and shows the level of auto-PEEP.

Causes of auto-PEEP include bronchospasm, blood or mucus within the large airways, or an expiratory time that is too short. High levels of auto-PEEP can cause dynamic lung hyperinflation and cardiovascular collapse. Treatment is described subsequently under the heading Bronchospasm and Obstructive Lung Disease.

Graphic Displays

Most ventilators can display real-time plots of derived respiratory indexes, which may be of assistance in adjusting ventilator settings.²⁴ The most widely used are pressure-time and flow-time curves, dynamic pressure-volume loops, and flow-volume loops.

Pressure-Time and Flow-Time Curves.

Pressure-time and flow-time curves (see Figs. 29-2 to 29-4) are referred to in the earlier sections on flow generation, end-inspiratory pressure, and end-expiratory pressure. They are helpful in diagnosing abnormalities in compliance, resistance, and the presence of auto-PEEP.

The flow-time curve identifies whether flow has fallen to zero at end-expiration and end-inspiration. Continued flow at end-expiration indicates the presence of auto-PEEP (see Fig. 29-4). This may be remedied by increasing the expiratory time, which is usually achieved by reducing the breath rate. Continued flow at end-inspiration with pressure-control ventilation (see Fig. 29-3) may indicate increased airway resistance. It may be remedied by increasing the inspiratory time. However, if auto-PEEP exists as well, this may not be possible, and higher airway pressures may be required.

Dynamic Pressure-Volume Loops.

Dynamic pressure-volume loops (Fig. 29-5) provide a graphic representation of airway pressure against volume in real time for each breath. A dynamic pressure-volume loop differs from a static compliance curve because, in addition to the effect of respiratory system compliance, the effects of airway resistance, circuit resistance, and airflow are included.²⁵

Figure 29.5 Pressure-volume loops. A, Two static pressure-volume curves (y and z; see also Fig. 1-15). Lower and upper inflexion points are shown on both curves. Above and below these inflexion points, compliance (ΔV/ΔP, the slope of the curve) is reduced. The lower inflexion point represents the lung volume at which some alveoli’s airways close (closing capacity); the upper inflexion point represents the start of overdistension. In curve y, the lower inflexion point lies below the functional residual capacity (FRC) and no airway closure occurs. In curve z, the inflexion point lies above the FRC, and airway closure will occur unless the FRC is increased by the application of PEEP. B, A dynamic pressure-volume curve is obtained by ventilating the lungs from residual volume (RV) to total lung capacity (TLC) when airways resistance is trivial. Hysteresis, in which lung volumes are different for a given pressure during inspiration and expiration is shown. Hysteresis is partly (but not solely) due to airway resistance. C, A dynamic pressure-volume curve is obtained when lungs are ventilated with a normal tidal volume from the FRC and airway resistance is normal. No inflexion points are seen, indicating no airway closure or overdistension. The shaded areas represent work to overcome respiratory system compliance (vertical hatch) and airway resistance (horizontal hatch). D, A dynamic pressure-volume curve with a lower inflexion point is indicative of inadequate PEEP. If a static pressure-volume curve were obtained, the FRC would be below the lower inflection point (curve z, A). E, A dynamic pressure-volume curve with an upper inflexion point is indicative of overinflation. This is termed “beaking.” F, The effect on the dynamic pressure-volume curve of increased airway resistance. Peak airway pressure (PAWP) and work of breathing (hatched areas) are increased because of the increased airway resistance. If expiration is active, the expiratory part of the curve will resemble the line labeled AE. Work done during active expiration is represented diagrammatically by the area labeled active exhalation work.

On a static respiratory system compliance curve (see Figs. 1-15 and 29-5A), upper and lower inflexion points can be identified. The upper inflexion point identifies overdistension of alveoli; the lower inflexion point identifies closing capacity. Ideally, the lungs should be ventilated between these two inflexion points, on the steep part of their compliance curves; that is, FRC should be above the lower inflexion point (as shown in Fig. 29-5A, curve y). The construction of a static compliance curve is impractical for routine care, and dynamic compliance curves are used instead (see Fig. 29-5B and C). Some ventilators can create an approximation of a static compliance curve by using the dynamic compliance curves generated by breaths of differing tidal volumes.

Inflexion points may be visible on dynamic-pressure volume curves but often they are not obvious. The presence of a lower inflexion point (see Fig. 29-5D) may indicate insufficient PEEP, whereas the presence of an upper inflexion point (see Fig. 29-5E) usually indicates overdistension due to excessive tidal volume. In patients ventilated appropriately with adequate PEEP and normal tidal volumes (6 to 8 ml/kg), ideally no inflexion points should be seen (see Fig. 29-5C). Changes in the dynamic pressure-volume loop that occur with increased airway resistance are shown in Figure 29-5F.

Flow-Volume Curve.

Normal and abnormal flowvolume loops are shown in Figure 29-6. Characteristic abnormalities may help to identify the presence of auto-PEEP, endotracheal tube cuff leak, or the accumulation of airway secretions.

Figure 29.6 Flow-volume loops. A normal flow-volume curve is shown along with the expected changes that occur with a, an endotracheal cuff leak, and b, dynamic lung hyperinflation (DLHI) and airway secretions.

Managing Impaired Gas Exchange

Ventilator Management

Increasing F_Io₂.

Increasing the F_Io₂ is only partially effective in treating hypoxemia due to V/Q mismatch. Prolonged periods of ventilation at an F_Io₂ above 0.6 may exacerbate lung injury and should be avoided if possible.²⁶ In patients with severe lung injury this may mean accepting a lower S_ao₂.

Altering the Tidal Volume.

To a certain extent, increasing tidal volume improves gas exchange, particularly carbon dioxide elimination. However, large tidal volumes (>10 ml/kg) overdistend open lung units and do not effectively recruit closed lung units, increasing the risk for ventilator-induced lung injury. Tidal volumes should be no higher than 10 ml/kg and preferably 6 to 8 ml/kg of ideal body weight.⁷

Decelerating Inspiratory Flow.

Changing from a constant inspiratory flow (volume control mode) to decelerating inspiratory flow (pressure control or hybrid volume control mode) may improve oxygenation and enhance carbon dioxide removal. However, there is no strong evidence to support the use of pressure control ventilation over volume control ventilation in patients with ARDS.²⁷

Altering Respiratory Rate.

With increased airflow resistance, decreasing the respiratory rate to 6 to 8 breaths per minute reduces the risk of air trapping and dynamic lung hyperinflation.²⁸ In restrictive lung disease (atelectasis, pulmonary edema, pneumonia, etc.), increasing the respiratory rate to 15 to 20 breaths per minute may improve gas exchange.

Increasing Inspiratory Time.

Moderate increases in inspiratory time (from 1.2 seconds to between 1.5 and 2 seconds) may improve oxygenation. This is caused mainly by the generation of auto-PEEP, but it may also be due to recruitment of lung units with high time constants. There is little additional oxygenation benefit with inverse ratio ventilation (e.g., I:E ratio 2:1 or 3:1), but if PEEP is already high, there is a substantial risk for causing dynamic lung hyperinflation due to excessive auto-PEEP.

Increasing PEEP.

Increasing PEEP (to 5 to 15 cm H₂O, or occasionally higher) is a first-line intervention for impaired oxygenation, but it should be used cautiously when there is right ventricular dysfunction or hypovolemia because it can cause marked hypotension.

Improving Patient-Ventilator Synchrony.

By optimizing ventilator settings and patient sedation, it is usually possible to have the patient breathing spontaneously with support from the ventilator. This optimizes V/Q matching and may make oxygenation better than it would be if the patient were paralyzed.

VENTILATORY STRATEGIES

Simple Adjuvant Therapies

Humidification

Humidification (see Chapter 28) of respiratory gases is essential to maintain mucociliary function and to prevent mucus plugging of large airways. Heat and moisture exchangers attached to the end of the endotracheal tube are suitable for use in patients who are mechanically ventilated for up to 48 hours.²⁹ For patients who are ventilated for longer than 48 hours, a heated water-bath humidifier should be used.

Tracheal Suctioning

Intermittent tracheal suctioning prevents mucus plugging and lobar collapse but should be performed only when clinically indicated by coarse or noisy breath sounds, increased airway pressures, or deteriorating gas exchange. Closed (sterile) suctioning systems do not confer a benefit in terms of reduced incidence of ventilatorassociated pneumonia.³⁰ Instillation of 0.9% saline (5 to 10 ml) and hand ventilation using a manual resuscitator is a useful maneuver if there is acute ventilatory compromise due to mucus plugging, but it should be performed only if normal suctioning is ineffective because it can have detrimental effects on oxygenation in the short term.³¹

Excessive suctioning may be harmful because of the loss of PEEP and the potential for airway trauma. This is particularly true for patients with ARDS and pulmonary edema. In patients who require high levels of PEEP, the use of an inline suctioning system minimizes the loss of PEEP during suctioning. A recruitment maneuver (see subsequent material) should follow suctioning in patients with severe lung injury who require high levels of PEEP.

Patient Positioning

To reduce the risk for pulmonary aspiration and ventilator-associated pneumonia, patients should be nursed in a 30- to 45-degree head-up position. In cases of unilateral pulmonary injury (e.g., pneumonia), nursing the patient in a semilateral position with the injured lung up maximizes blood flow to the uninjured lung and usually improves oxygenation.

Other Therapies

Oxygenation is improved by avoiding drugs that inhibit hypoxic pulmonary vasoconstriction and by increasing low mixed-venous oxygen saturation (S_vo₂). Any systemically administered pulmonary vasodilator, such as nitroglycerine, milrinone, and nitroprusside, will inhibit hypoxic pulmonary vasoconstriction. Low S_vo₂ can be improved by treating low cardiac output and anemia and by reducing metabolic rate. Metabolic rate is reduced by sedation, paralysis, avoiding overfeeding, and active cooling (with paralysis).

Advanced Adjuvant Therapies

Recruitment Maneuvers

Recruitment maneuvers may be helpful in reopening collapsed alveoli.³² A recruitment maneuver involves the application of 30 to 40 cm H₂O PEEP for 30 to 40 seconds. During the maneuver, patients should be heavily sedated or paralyzed so as to avoid respiratory effort. The maneuver may need to be terminated early if the patient develops severe hypotension or hypoxemia. Following the maneuver, PEEP is set to a very high level (20 to 25 cm H₂O) and reduced in 2.5-cm-H₂O increments every 10 minutes while observing the effect on gas exchange. The level of PEEP at which oxygenation deteriorates is the point at which airways closure is occurring. The recruitment maneuver should then be repeated and PEEP applied at a level just above that at which airway closure occurs. If no benefit is observed as a result of the intervention, no further recruitment maneuvers should be performed.³³

A recruitment maneuver may be considered in a patient who requires an F_Io₂ greater than 60% and PEEP greater than 10 cm H₂O for a sustained period. Benefit from a recruitment maneuver is most likely in patients with atelectasis and ARDS. It is more effective in the prone position. In the setting of hypovolemia or ventricular dysfunction, recruitment maneuvers can cause severe hypotension and must be performed with extreme caution.

Prone Positioning

Atelectasis is most likely to affect dependent pulmonary segments (see Fig. 27-2) which, because of the effects of gravity, are also the regions that receive preferential blood flow. With prone positioning, atelectatic regions are predominantly nondependent, and dependent regions are well ventilated and well perfused, which reduces intrapulmonary shunting. Prone positioning for 6 hours has been shown to improve oxygenation in 70% of patients with ARDS or acute lung injury (ALI),³⁴ but a survival benefit has not been clearly demonstrated.³⁵ Patients with ARDS of nonpulmonary origin (e.g., sepsis) obtain the greatest benefit. Prone positioning for 6- to 12-hour periods, alternating with 6- to 12-hour periods of supine positioning, may be considered in patients who require an F_Io₂ above 60% and PEEP above 10 cm H₂O for a sustained period. If benefits are obtained, it should be continued until the F_Io₂ is less than 40% to 50% and PEEP less than 10 cm H₂O.

In the prone position, the patient must be supported, e.g., by pillows, under the chest and pelvis, leaving the abdomen free for respiratory excursion. The patient’s head should be supported primarily under the forehead. The eyes, nose, lips, airway, breasts, and genitals must be carefully inspected for pressure points or compression. Lifting and turning large adult patients require a coordinated effort by at least five people.

Tracheal Gas Insufflation

A small-bore catheter may be placed through or beside the tracheal tube, with the end in the lower portion of the trachea and a low flow of oxygen administered (e.g., 2 l/min) to increase F_Io₂, augment PEEP, and wash out carbon dioxide.³⁶

Inhaled Nitric Oxide

Inhaled nitric oxide is a potent pulmonary vasodilator that is used to treat severe pulmonary hypertension and right ventricular failure (see Chapter 24).³⁷ Because inhaled nitric oxide is delivered only to ventilated regions of the lung, it also improves V/Q matching within the lung. Doses of 5 to 40 parts per million improve oxygenation in some patients with ARDS. However, the benefit is sustained for only 24 to 48 hours, and no improvement in patient outcome has been demonstrated.³⁸ Inhaled nitric oxide may be considered in patients with severely impaired gas exchange that is resistant to other ventilatory measures. Sudden cessation of nitric oxide can cause marked hypoxemia and increased pulmonary vascular resistance, which can precipitate right ventricular failure.

Corticosteroids

With the exception of bronchospasm, corticosteroids have no role in the management of acute respiratory failure. There are conflicting data concerning the role of corticosteroids in persistent ARDS. One small, randomized trial involving only 24 patients demonstrated a marked survival benefit in patients with severe ARDS who had failed to improve by the seventh day of respiratory failure.³⁹ However, a recent randomized trial involving 180 patients did not demonstrate any difference in survival rates in patients with ARDS of at least 7 days’ duration between those who were treated with methylprednisolone (2 mg/kg then 0.5 mg/kg every 6 hours for 14 days with subsequent tapering of the dosage) and those who were treated with placebo.⁴⁰ Also, mortality rates were higher in the treatment group when methylprednisolone was started 2 or more weeks after the onset of ARDS. Some secondary outcomes, such as measures of gas exchange and number of ventilator days, were, however, better in steroid-treated patients. Routine use of corticosteroids for ARDS is not recommended.⁴¹

Specific Clinical Scenarios

Hypoxemia in the Early Postoperative Period

Occasionally, patients arrive from the operating room in a state of hypoxemia or hypoxemia rapidly develops in the ICU. Initial management is to rule out a life-threatening or easily correctable cause. The patient (along with the ventilator and breathing circuit) should be examined, and the postoperative chest radiograph should be reviewed. The common causes of hypoxemia in this situation are listed in Table 27-1.

Sometimes no cause for the hypoxemia is identified: the lungs are easy to ventilate, the chest radiograph is unremarkable, and P_aco₂ and pH are normal. This situation usually represents a combination of the adverse effects of positive pressure ventilation on V/Q matching, microatelectasis, impaired hypoxic pulmonary vasoconstriction, increased extravascular lung water, and low S_vo₂. It may be useful to suction the endotracheal tube and provide a gentle recruitment maneuver. Depending on the severity of the hypoxemia, no specific treatment may be required and the problem usually resolves over a few hours. If intervention is required, the following should be considered: (1) increasing PEEP; (2) stopping drugs that impair hypoxic pulmonary vasoconstriction; (3) using a decelerating inspiratory flow pattern; (4) increasing inspiratory time to more than 1.5 seconds; (5) deep sedation and paralysis. Superior vena cava oxygen saturation (S_SVCo₂) should be measured. Pulmonary artery catheterization, an echocardiogram, or both should be performed if there is hemodynamic instability or if S_SVCo₂ is low (see Chapter 20).

Patient-Ventilator Dysynchrony

Patient-ventilator dysynchrony is common when patients are awakened after cardiac surgery. Disorientation, biting the endotracheal tube, breathholding, shivering, and inappropriate ventilator settings all contribute to dysynchrony. It can present with impaired gas exchange, pulsus paradoxus, hypotension, and elevated central venous pressure. The hemodynamic picture may be identical to that of cardiac tamponade. The patient usually appears distressed, but that is not always the case.

If the patient is unable to trigger the breaths, the trigger threshold should be reduced. If the ventilator is autotriggering, the trigger threshold may have to be increased. Other ventilator settings that may require adjustment include the ventilation mode, the method of cycling, and the inspiratory rise time. If ventilator dysynchrony occurs as the patient is being awakened, verbal reassurance may be effective. However, if the patient is disorientated and uncooperative, resedation with or without paralysis may be required. Once analgesia, anxiolysis, and ventilator settings have been optimized, the patient should be awakened again.

Obesity

Patients who are morbidly obese are more likely to develop hypoxemia and require high airway pressures when mechanically ventilated. With obesity, FRC is reduced and closing capacity is elevated, resulting in increased V/Q mismatch (see Chapter 1). Chest wall compliance is reduced because of the weight of the thorax, abdomen, and breasts.

Patients should be nursed in a 30- to 45-degree head-up position. A strategy of small tidal volumes (6 ml/kg of ideal body weight) and a relatively rapid breath rate (14 to 18 breaths per minute) using pressure control ventilation helps to keep airway pressures to a minimum. Modest hypercapnia (pH >7.25) is acceptable. Despite this, increased peak and plateau airway pressures are common. However, high airway pressures are not especially dangerous because usually the transalveolar pressure is not increased (see Fig. 29-1).

Bronchospasm and Obstructive Lung Disease

Severe bronchospasm is a medical emergency. However, before a patient is treated for bronchospasm, other causes of ventilatory difficulty should be ruled out, particularly heart failure, mechanical airway obstruction, and pneumothorax. Bronchospasm may be precipitated by tracheal suctioning, by the lightening of sedation, or because the tip of the endotracheal tube has touched the carina.

The two major risks involved in invasive ventilation in patients with severe bronchospasm are (1) cardiovascular collapse secondary to dynamic lung hyperinflation and (2) tension pneumothorax. Dynamic lung hyperinflation may be avoided by ventilating with low breath rates and long expiratory times. Ventilation with a low minute volume (accepting a high P_aco₂) and maintenance of a normal inspiratory time minimize peak airway pressures. High peak airway pressures in this setting are not indicative of high transalveolar pressures (see Fig. 29-1); peak airway pressures above 35 cm H₂O are preferable to shortening the expiratory time because the risk for barotrauma arises mainly from dynamic pulmonary hyperinflation or rupture of a bulla.

The following initial ventilator settings are appropriate for severe bronchospasm:

• Volume control ventilation with a tidal volume of 6 ml/kg or pressure control ventilation with a peak pressure of 30 to 35 cm H₂O

• Minimal PEEP (An initial level of 0 to 5 cm H₂O is appropriate, but it may be set to match the level of auto-PEEP.)

• Breath rate of 6 to 8 per minute, with an inspiratory time of 1.3 seconds

• Pressure support of 20 cm H₂O

• F_Io₂ of 100%.

Bronchodilators may be administered into the ventilator circuit by nebulization or metered dose inhaler (see Chapter 28). In cases of hemodynamic collapse in which dynamic lung hyperinflation is suspected, the patient should be disconnected from the ventilator for 30 to 40 seconds to allow full expiration to occur (the “Lazarus maneuver”).

Cardiac Failure

Positive pressure ventilation reduces left ventricular afterload (see Chapter 1) and decreases extravascular lung water (pulmonary edema), which is beneficial in patients with left ventricular dysfunction. In the first instance, patients may respond to noninvasive ventilation (see Chapter 28). If invasive ventilation is required, no specific mode is recommended but high PEEP (>10 cm H₂O) is rarely required. With pulmonary edema, disconnections from the ventilator to allow tracheal suctioning should be kept to an absolute minimum. Not only is suctioning ineffective in controlling pulmonary edema, but disconnecting the ventilator results in the loss of PEEP, which allows further extravasation of edema fluid.

The effects of positive pressure ventilation on right ventricular function are variable. High intrathoracic pressure may cause marked hypotension due to reduced right ventricular preload and increased right ventricular afterload. However, if positive pressure ventilation improves gas exchange and maintains FRC, right ventricular function may improve because of a fall in right ventricular afterload.

ALI/ARDS and Lung-Protective Ventilation

Severe ARDS is relatively rare in cardiac surgery patients, but it may be encountered more frequently if the ICU functions as a referral center for extracorporeal membrane oxygenation (ECMO). ARDS is characterized by reduced lung compliance, impaired gas exchange, and a high risk for barotrauma (see Chapter 27). The goals in ventilating patients with ARDS are to sustain life while lung function recovers and to avoid ventilator-induced lung injury. The following ventilator settings and physiologic goals are appropriate:

• F_Io₂. This should be set to achieve an S_ao₂ above 90%. However, if this requires an F_Io₂ greater than 60%, the goal should be reduced to an S_ao₂ above 85%.

• Tidal volume. This should be 6 ml/kg but should be further reduced if plateau pressure is greater than 32 cm H₂O.

• Respiratory rate. A rate of 15 to 20 breaths per minute is usually appropriate. The breath rate (or inspiratory time) may have to be reduced if auto-PEEP develops and may have to be increased if there is profound acidosis (see subsequent material).

• Inspiratory flow pattern. A decelerating flow pattern, with either pressure control or hybrid flow generation ventilation, should be used.

• Inspiratory time. This should be 1.3 to 2 seconds. Longer inspiratory times may be tried in patients with profound hypoxemia.

• PEEP. This should be 10 to 20 cm H₂O, depending on oxygenation, the response to recruitment maneuvers, and the shape of the dynamic pressure-volume loop (see Fig. 29-5D).

• pH and P_aco₂. This ventilatory strategy is likely to result in hypercarbia and respiratory acidosis. However, there is no evidence that even severe hypercarbia causes significant morbidity and, in fact, it may be beneficial.⁴² Whereas a pH less than 7.05 to 7.1 should probably be avoided, a pH of 7.2 is preferable to high airway pressures. With normal renal function, metabolic compensation for a respiratory acidosis develops over a few days, allowing a very high P_aco₂ (>10 κPa or 76 mmHg) to be associated with a pH above 7.2.

The adjuvant ventilatory and nonventilatory strategies outlined should be considered, particularly recruitment maneuvers, prone positioning, inhaled nitric oxide, and reduction of metabolic rate. The strategy of small tidal volume ventilation with permissive hypercarbia is known as lung-protective ventilation, and it improves survival rates in patients with ARDS.⁷ For patients with life-threatening respiratory failure despite the treatment outlined, high-frequency oscillation or venovenous ECMO should be considered.

High-Frequency Oscillation

Patients with potentially reversible acute respiratory failure are occasionally unable to be kept alive with conventional ventilation. This most commonly occurs following lung transplantation and in patients with severe pneumonia complicated by ARDS. Such patients may be considered for either high-frequency oscillation or venovenous ECMO (see Chapter 22). High-frequency oscillation should be considered prior to ECMO in patients who fulfill ECMO criteria (see Chapter 22) but do not have severe hemodynamic instability.

A high-frequency oscillator contains a piston that moves back and forth rapidly as fresh gas (the bias gas flow) passes in front of it. The ventilator generates a constant positive airway pressure onto which small tidal breaths (<2 ml/kg) at very high rates (300 to 900/min) are superimposed.⁴³ The constant positive airway pressure helps to recruit closed alveoli and therefore improves oxygenation, and the rapid, small breaths (oscillations) remove carbon dioxide.⁴⁴

Patients may deteriorate for a short time when first placed on the oscillator and may take several hours to show improvement. The benefit of high-frequency oscillation is lost if there are frequent disconnections from the ventilator for suctioning. The following are typical adult settings:

• Mean airway pressure. This should be set at 2 to 4 cm H₂O above that which the patient was receiving with conventional ventilation. One hour after commencing high-frequency oscillation, a chest radiograph should be obtained and lung expansion assessed. Between 9 and 10 ribs should be visible posteriorly. If fewer are visible, mean airway pressure should be increased; if more are visible, airway pressure should be decreased. Mean airway pressure may also be increased to improve oxygenation.

• Breath rate. This should be set to between 300 and 480 breaths per minute (5 to 8 Hz).

• Power (amplitude). Power determines the movement of the piston. It should be set so that the chest visibly vibrates, generating a respiratory pressure amplitude that is typically 30 to 60 cm H₂O.

• Bias flow. This should be set to 40 to 60 l/min.

• Inspiratory time. This is typically set to 33%.

• F_Io₂. This should initially be set to 100%.

Oxygenation is determined largely by F_Io₂ and mean airway pressure, and these should be altered if P_ao₂ is unacceptably low. Carbon dioxide elimination is determined mainly by breath rate and to a lesser extent by bias flow and power. If hypercarbia and acidosis are severe, breath rate (frequency) should be decreased and power (amplitude) may be increased. If the patient has severe respiratory acidosis despite these maneuvers, deflating the cuff on the endotracheal tube to allow a small air leak may result in improved elimination of carbon dioxide.

Once the patient has been established on highfrequency oscillation, he or she should be weaned to an F_Io₂ of 0.6. Next, mean airway pressure should be reduced in steps of 2 cm H₂O. This should be performed slowly so as to avoid airway closure. Once mean airway pressure is 10 to 14 cm H₂O, the patient may be converted to conventional ventilation.

Ventilator Alarms

The causes and treatment of commonly employed alarms are listed in Table 29-1.

Table 29-1 Ventilator Alarms

Alarm	Possible Causes	Interventions
High airway pressure with VC or low tidal volume with PC	Mucus plugging	Suction ETT
		Consider saline lavage
	Blocked circuit or HME	Ventilate with manual resuscitator, change HME and circuit
	Endobronchial intubation	Review CXR and withdraw ETT
	Obesity	Have patient sit up
		Accept higher airway pressure
	Patient-ventilator dysynchrony	Reassure patient
		Adjust ventilator settings (see text)
		Sedate and/or paralyze
	Increased airway resistance ± dynamic lung hyperinflation	Examine patient and ventilator graphics to confirm increased resistance ± auto-PEEP
		Reduce breath rate and ensure adequate I_t and E_t
		Treat bronchospasm if present
		Treat bronchospasm if present
		Suction if mucus plugging exists
		Decrease set PEEP
		In emergency: disconnect patient from ventilator for 30 seconds
	New lung disease causing reduced compliance (e.g., pulmonary edema, hemothorax, pneumothorax)	Examine patient and check ventilator graphics (see text)
		Review CXR
		Treat specific pathology (see text)
	Blocked expiratory valve	Disconnect patient from ventilator
		Hand ventilate with manual resuscitator
Low expired minute volume	Spontaneous mode: apnea or hypoventilation	Change to mandatory ventilation mode
	Mandatory mode: cuff leak, circuit disconnection, large air leak from lung	Examine patient, circuit, and ventilator for leak
		Check chest drains for bubbling; if present, consider reducing suction/lung isolation/surgery
High expired minute volume	Hyperventilation, autotriggering, excessive pressure support or mandatory ventilation	Examine patient
		Ensure patient is not developing metabolic acidosis
		Check that ventilator settings are appropriate
		Treat respiratory alkalosis as appropriate (see text for details)
High respiratory rate	Autotriggering: Water in the ventilator circuit, patient movement, hyperdynamic cardiac impulse, large cuff leak	Remove water from circuit
		Sedate patient
		Inflate cuff
	Inadequate respiratory support from ventilator	Decrease trigger sensitivity
		Increase trigger sensitivity
		Increase ventilatory support
	Inadequate sedation	Optimize sedation

CXR, chest radiograph; E_t, expiratory time; ETT, endotracheal tube; HME, heat and moisture exchanger; I_t, inspiratory time; PC, pressure control; VC, volume control; PEEP, positive end expiratory pressure.

Weaning From Ventilation

For routine cardiac surgery patients, weaning from invasive ventilation is not required. Rather, sedation is stopped and, once patients begin making respiratory efforts, they are changed to a spontaneous breathing mode and extubated when awake (see Chapter 17).

Prolonged invasive ventilation for serious respiratory or cardiac disease involves two phases: (1) rest and repair; (2) recovery and retraining. During rest and repair, the aim is to reduce the work of breathing, provide adequate gas exchange, and avoid ventilator-induced lung injury. Patients are typically sedated but routine neuromuscular paralysis is not indicated because it prolongs the duration of invasive ventilation and increases the risk for critical illness polymyoneuropathy. Assist/control and SIMV with pressure support are commonly used modes during the rest and repair phase. Weaning (the recovery and retraining phase) is commenced once certain criteria have been met (Table 29-2).⁴⁵

Table 29-2 Criteria for Commencing Weaning of Ventilatory Support

Rights were not granted to include this table in electronic media. Please refer to the printed book.

From MacIntyre NR, Cook DJ, Ely EW Jr, et al: Evidence-based guidelines for weaning and discontinuing ventilatory support: a collective task force facilitatd by the American College of Chest Physicians, the American Association for Respiratory Care, and the American College of Critical Care Medicine. Chest 120:375S-395S, 2001. PEEP, positive end-expiratory pressure.

Methods of Weaning

Multiple ventilatory techniques have been used to facilitate weaning. They include: (1) assist/control mode; (2) SIMV with a gradual reduction in the breath rate; (3) pressure support mode with a gradual reduction in the level of support; (4) frequent trials of spontaneous breathing. Trials of spontaneous breathing or a gradual reduction in the level of pressure support result in shorter periods of ventilation than does an SIMV-based weaning strategy.⁴⁶ In one study, trials of spontaneous breathing were superior to gradual reductions in pressure support.⁴⁷

Regardless of the weaning technique, the following should be borne in mind. First, the use of a protocol shortens weaning time.⁴⁸ Second, the time required for weaning varies considerably among patients, from a few hours to longer than a month. Thus, clinicians must be prepared to modify weaning plans according to individual patients’ responses. Third, patients should never be allowed to exhaust themselves because it causes a loss of muscle strength that may require a further period of rest and recovery before weaning can be attempted again.

Trials of Spontaneous Breathing

A trial of spontaneous breathing involves disconnecting the patient from the ventilator and allowing him or her to breathe through a T-tube or CPAP circuit (with 5 cm H₂O of positive pressure). The frequency and length of trials are dictated by the patient’s response, but a trial should be done at least once a day. Initially, only 30 to 60 minutes may be managed. Between trials (and overnight), full ventilation is provided. Patients may be extubated once they have tolerated several hours of spontaneous ventilation without respiratory distress.

Pressure Support Weaning

Pressure support is commenced at a level of 15 to 20 cm H₂O (above PEEP) and then adjusted to maintain a breath rate of fewer than 25 to 35 breaths per minute. The level of pressure support is reduced twice daily in steps of 2 to 4 cm H₂O and is increased only if the patient develops respiratory distress. In a patient who is very comfortable, the frequency of pressure reduction may be increased. A PEEP of at least 5 cm H₂O should be maintained throughout. The patient is extubated once a pressure support level of 5 to 10 cm H₂O above PEEP has been reached.

Monitoring During Weaning

Oxygen saturation, breath rate, tidal volume (if available), and heart rate should be monitored continuously during weaning. A breath rate greater than 30 to 35 breaths per minute, a tidal volume less than 3 to 4 ml/kg, tachycardia, sweating, use of accessory muscles, and intercostal indrawing indicate impending exhaustion.

Pulse oximetry is used to guide F_Io₂ and PEEP settings. Measurements of P_aco₂ and end-tidal carbon dioxide concentration are of little help because they usually increase only once the patient has become overtired. Thus regular measurement of arterial blood gases is not necessary during weaning. Chest radiography should be performed only when clinically indicated.

Failure to Wean

Any correctable respiratory problems should be identified and treated. This may require diagnostic chest ultrasound, chest computed tomography, diaphragmatic screening, and bronchoscopy. Pneumothoraces and significant pleural collections (>300 to 400 ml on ultrasound) should be drained. Pulmonary edema should be treated with diuretics. Pneumonia should be actively sought and treated with appropriate antibiotics. Lobar collapse and basal atelectasis may be treated with bronchoscopy-guided suctioning and chest physical therapy, which may include postural drainage and incentive spirometry. Treatment of phrenic nerve injury is usually supportive but occasionally diaphragmatic plication is required.

Cardiac dysfunction is a common cause of failure to wean.⁴⁹ A patients who is difficult to wean should undergo an echocardiogram. If left ventricular systolic function is impaired, treatment with angiotensin-converting enzyme inhibitors, digoxin, furosemide, and spironolactone may facilitate weaning. Significant pericardial collections should be drained.

Excessive fluid administration can increase extravascular lung water and impair respiratory function. Also, patients receiving positive pressure ventilation tend to retain sodium and water. Daily water requirements are low in ventilated patients—about 20 ml/kg/day (see Chapter 32). The daily fluid balance should be closely scrutinized, and treatment with furosemide may be warranted even if cardiac function is normal.

Optimal nutrition is vital during weaning. Excess calories,⁵⁰ particularly in the form of carbohydrates,⁵¹ increase carbon dioxide production and increase respiratory work. Too few calories prolong weakness and contribute to ventilator dependence.

Critical illness polymyoneuropathy (see Chapter 37) is a common cause of prolonged weaning. Treatment is supportive, and exacerbating factors, such as the use of corticosteroids and neuromuscular blockers, should be kept to a minimum.

Patients who have suffered a neurologic insult as a result of cardiac surgery may be difficult to wean because of weakness, agitation, or depressed consciousness.

Acid-Base Disturbance During Weaning

Acid-base disturbances can accompany and complicate weaning. Metabolic alkalosis is common in long-stay cardiac surgery patients as the result of chronic use of diuretics and of fluid restriction. Treatment, if required, consists of replacing potassium losses and restoring intravascular volume with a 0.9% saline-containing solution.

Respiratory alkalosis is reasonably common during chronic weaning and may develop because of anxiety, pain, inappropriate ventilator settings, or increased respiratory drive. Increased respiratory drive may occur due to hypoxemia (P_ao₂ <8 κPa or 60 mmHg), neurologic injury, sepsis, or pulmonary embolus. Occasionally, in patients with neurologic injury, severe hyperventilation results in a dangerous increase in pH (>7.55). In this situation, intravenous hydrochloric acid or acetazolamide (250 to 500 mg, 2 to 4 times daily) may be given.

Respiratory acidosis may be caused by inadequate ventilatory support or high metabolic rate. If permissive hypercarbia has been employed as part of lung-protective ventilation, metabolic compensation may have occurred, normalizing pH. In this situation, normal pH, not normal P_aco₂, should be targeted during weaning. If hypercarbia is due to high metabolic rate secondary to sepsis, weaning should not be attempted until the metabolic state has normalized. Similarly, weaning should not be commenced if there is persistent metabolic acidosis.

Extubation

Numerous indexes have been proposed for predicting successful extubation (Table 29-3), but none have high positive predictive values.⁵² Prior to extubation, patients should be breathing comfortably and have satisfactory respiratory rates, tidal volumes, oxygenation (P_ao₂/F_Io₂ >20 κPa or 150 mmHg), and cough. They should be able to obey simple commands.

Table 29-3 Predictors of Successful Extubation

Rights were not granted to include this table in electronic media. Please refer to the printed book.

From MacIntyre NR, Cook DJ, Ely EW Jr, et al: Evidence-based guidelines for weaning and discontinuing ventilatory support: a collective task force facilitated by the American College of Chest Physicians; the American Association for Respiratory Care; and the American College of Critical Care Medicine. Chest 120:375S-395S, 2001.

Tracheotomy

Every patient who is ventilated for longer than 7 to 10 days should be considered for a tracheotomy, ideally at the beginning of the recovery and retraining phase.⁵³ Early tracheotomy (within the first week) may reduce the total number of days on the ventilator.⁵⁴ A tracheotomy provides greatly increased tolerance of ventilation, more effective clearance of secretions, the ability to remove and reinstitute invasive ventilation easily, and the potential to speak and eat.⁴⁵ The main disadvantages of tracheotomies are procedure related. Tracheotomies do not increase the incidence of sternal wound infection or mediastinitis in cardiac surgery patients.⁵⁵^–⁵⁷ Percutaneous insertion is at least as safe as open surgical placement⁵⁸ and has the advantage of being able to be performed at the bedside. The technique of percutaneous tracheotomy is described in Chapter 40.

Management of Tracheotomies

A large-sized tracheotomy tube—typically 9 mm internal diameter for males and 8 mm internal diameter for females—should be used so as to minimize airway resistance. Initially, a nonfenestrated tube with a high-volume, low-pressure cuff should be used. The cuff should be inflated during positive pressure ventilation but may be deflated during periods of T-tube breathing. An inner tube may be used; it has the advantage of being able to be removed and cleaned. Adequate humidification of inspired gas should be provided by a heated water-bath (see Chapter 28); heat and moisture exchangers are suitable only for short-term use, and they impose increased circuit resistance.

In the case of a correctly placed tube of appropriate size, no airway leak should occur with a cuff pressure of less than 20 cm H₂O and a volume of less than 10 ml. Cuff pressure should be checked regularly with a manometer. The presence of an airway leak despite appropriate cuff volume and pressure suggests that the tracheotomy tube has become partially dislodged (see Chapter 40).

With percutaneous insertion, routine changes of the tracheotomy tube are not recommended, particularly in the first few days, because the tube is likely to be tightfitting and the track not well formed. If the tube becomes dislodged or blocked, it may be impossible to reposition it, even with aid of a bronchoscope. Facilities for orotracheal intubation should be immediately available during any manipulations of a tracheotomy tube.

Speaking.

When a patient is breathing spontaneously on a T-tube with the cuff down, a one-way (inhalation only) valve may be attached to the tracheotomy tube to allow the patient to exhale around the tube and through the vocal cords. In this way, speech is possible. Speech may be facilitated by the insertion of a fenestrated tracheotomy tube (Fig. 29-7A and B), which can be used either without an inner tube or with a fenestrated inner tube is a patient with an established tracheotomy. However, this requires changing the tracheotomy tube, which should be performed no sooner than 1 week after percutaneous insertion, when the tracheotomy track is well formed. Although positive pressure breathing is technically possible with a fenestrated tracheotomy combined with a nonfenestrated inner tube (Fig 29-7 D), it may be problematic due to airway leak, and a fenestrated tube should be inserted only once the patient has been weaned, for the most part, from positive pressure ventilation.

Figure 29.7 A fenestrated tracheotomy tube with fenestrated and nonfenestrated inner tubes.

Swallowing and Eating.

Between 30% and 50% of patients with tracheotomies are observed to aspirate when swallowing is examined by video fluoroscopy, irrespective of whether or not the cuff is inflated. In 80% of cases, aspiration is clinically silent.^59,⁶⁰ During swallowing, a tracheotomy tube prevents normal laryngeal elevation, which is one of the mechanisms of protecting the airway from soiling by food particles. The inflated tracheotomy tube cuff also alters upper esophageal function and predisposes to aspiration. Patients with tracheotomies are often weak, with poor cough and impaired vocal cord function. Thus, patients should not routinely be fed orally; rather, nutrition should be provided via a fine-bore nasogastric or nasojejunal tube. However, during prolonged weaning, eating may significantly enhance a patient’s mood and oral comfort and the desire to get better. If a decision is made to offer a trial of feeding, the patient should be helped into a sitting position, the cuff deflated, and a small volume of soft food offered; soft food is less likely to be aspirated than is a liquid.