7: Mechanical Ventilation

Ventilator management for most thoracic surgery patients involves two distinct phases: (1) support in the operating room while the patient is undergoing surgery and receiving general anesthesia, and (2) support in the postoperative recovery room or intensive care unit (ICU) as the patient is recovering from surgery. Issues relating to the intraoperative ventilator management of thoracic surgery patients are largely the responsibility of the anesthesiologist and are discussed in Chapter 5. Issues relating to the postoperative management of mechanical ventilation are the responsibilities of the thoracic surgeon and intensivist.

In the transition from the operating room to the recovery room or ICU it is important to appreciate that general anesthesia and thoracic surgery adversely affect nearly all aspects of respiratory physiology. These include anesthetic-related alterations in respiratory drive, reductions in lung volume due to loss of chest wall tone, changes in the ventilation perfusion relationship, and increased airway resistance in the setting of diminished lung volumes. Given the postoperative structural changes in the lung as well as its native state of disease, these alterations may have variable effect on overall function and be unpredictable in duration. They certainly must be taken into account during initial ventilatory management.

The overall approach to mechanical ventilation in the postoperative thoracic surgery patient is similar to that used in the critically ill medical patient. Preexisting lung disease, intraoperative complications, and known physiologic alterations associated with a planned surgery require more innovative approaches.

There are two basic approaches to mechanical ventilation in patients who have undergone thoracic surgery. These are (1) methods used to support postoperative patients who are kept intubated after surgery for a specific indication that is expected to resolve within hours, allowing for rapid discontinuation of ventilator support; and (2) methods used to support patients who develop hypoxic or hypercarbic respiratory failure as a consequence of a primary process that will resolve over a period of days to weeks and may require more gradual weaning.

In most patients, the physiologic alterations caused by anesthesia and thoracic surgery are well tolerated. These patients generally have minimal to mild preexisting pulmonary disease and are either extubated in the operating room or arrive in the postoperative recovery area or ICU ready for extubation with normal Pao₂ and Paco₂ blood gas values on minimal ventilator support. Successful extubation in this group is associated with: (1) intact mental status; (2) reasonable assurance that the patient will have the ability to cough and protect his or her airway; and (3) initiation of an analgesic protocol that optimizes respiratory mechanics without causing undue respiratory depression.

Although mental status is usually simple to assess, often it is not possible to confirm intact recurrent laryngeal nerve (RLN) function before attempting extubation simply by relying on the patient's ability to cough and swallow secretions. The risk of injury to the RLN is increased in the thoracic surgery population because many procedures involve anatomic dissection or traction on structures near the left mainstem bronchus where the RLN branches from the vagus.¹ Postextubation evaluation revealing a weak voice and ineffective cough should prompt direct laryngoscopic evaluation of the hypopharynx and vocal cords, followed by vocal cord medialization if indicated.²

Several factors contribute to respiratory muscle dysfunction after thoracic surgery. Pain is a major contributor. Thus, selection of an appropriate analgesic regimen is essential for preventing postoperative respiratory failure.³ Studies of respiratory muscle function also have demonstrated that diaphragmatic contractility is compromised by somatic reflex inhibition of the phrenic nerve as a consequence of afferent intercostal stimulation.⁴ Thus analgesic regimens that address both these factors should provide optimal management. Epidural anesthesia with local anesthetic agents (i.e., bupivacaine) accomplishes this objective and can be administered either alone or in combination with opioids. Patient-controlled anesthesia (PCA) also may be an effective adjunct, but care must be taken in titration to minimize the degree of respiratory depression.⁵

A substantial number of patients undergoing thoracic surgery will require ventilator support postoperatively. A successful ventilator strategy for longer-term management involves the following general principles:

1. Selecting a mode of ventilation that prevents high airway pressures and optimizes patient-ventilator synchrony.

2. Early weaning of Fio₂ to prevent adsorption atelectasis and limit possible oxygen toxicity, especially in patients receiving medications that have been associated with free-radical lung injury (e.g., bleomycin).

3. Selecting an appropriate sedation/analgesia regimen to ensure patient comfort while permitting periodic assessments of mental status and respiratory function.

4. Initiating nutritional support and deep venous thrombosis prophylaxis.

5. Closely monitoring the intravascular fluid status to prevent development of pulmonary edema, especially in areas of lung tissue that have been manipulated during surgery.

No guidelines exist for selecting the “single best” mode of ventilation for the postoperative thoracic surgery patient. Anecdotal experience indicates that many surgeons select pressure-controlled modes of ventilation in which the user-specified independent variable is airway pressure rather than tidal volume. This ensures that airway pressures will not exceed a known value, thus limiting stress on newly created staple or suture lines. There are no data to indicate that this approach improves respiratory physiology or ICU outcomes in this patient population.⁶ Furthermore, appropriate selection of ventilator parameters using volume-cycled modes can ensure equivalent limiting of airway pressures. Thus, in most instances, user preference and experience will dictate ventilator settings.

Mechanical ventilation has evolved dramatically in the last 100 years, revolutionizing the concept of intensive care. Mechanical ventilation can be provided as either negative pressure ventilation (iron lung, cuirass shell, or rocking bed) or more commonly as positive pressure inflation by using an endotracheal tube or mask (known as noninvasive positive pressure ventilation). Positive pressure ventilation is also classified as manual (hand bag valve mask) or mechanical. The latter is provided by transport, critical care, and neonatal/pediatric ventilators.

There are many models of commercially available ventilators and it is important for the physician to understand the core components of these devices (Fig. 7-1). The most widely used system is positive pressure ventilation, which is administered through an endotracheal tube.⁷ Positive pressure ventilation requires the following components: an oxygen source (if O₂ mixture above 21% is needed), a mixing chamber, a compressor, an electronically controlled inspiratory valve, a ventilator circuit (with inhalation, “Y” connector, and exhalation limb), and an exhalation valve to form a closed circuit. The core of the apparatus is a computer, which controls the manner in which the ventilator breaths are triggered, cycled, and limited. These parameters are organized as preset modes. The computer also monitors pressure and flow sensors, providing a feedback and safety system. The array of information generated by the computer is displayed on a monitor in tabular or graphic format. In this way, the clinician can adjust the patient's ventilatory strategy based on the patient's clinical status while minimizing potential harm.

Commonly used modes available on most commercial ventilators include assist/control, pressure-controlled ventilation (PCV), pressure-support ventilation (PSV), and continuous positive airway pressure (CPAP). The setup and operation of these modes are described below.

When choosing ventilator settings, the mode refers specifically to the manner in which ventilator breaths are triggered, cycled, and limited. The trigger, either an inspiratory effort or a time-based signal, defines what the ventilator senses to initiate an assisted breath. Cycle refers to the factors that determine the end of inspiration. For example, in volume-cycled ventilation, inspiration ends when a specific tidal volume is delivered to the patient. Other types of cycling include pressure cycling, time cycling, and flow cycling. Limiting factors are operator-specified values, such as airway pressure, that are monitored by transducers internal to the ventilator circuit throughout the respiratory cycle. If the specified values are exceeded, inspiratory flow is immediately stopped, and the ventilator circuit is vented to atmospheric pressure or the specified positive end-expiratory pressure (PEEP). A list of types of assessments that should be considered for a patient undergoing mechanical ventilation are provided in Table 7-1.

In assist/control mode ventilation (ACMV), an inspiratory cycle is initiated either by the patient's breathing effort or, if no patient effort is detected within a specified time window, by a timer signal within the ventilator based on user-specified parameters. Every breath delivered, whether patient- or timer-triggered, consists of the operator-specified tidal volume. Ventilatory rate is determined either by the patient or by the operator-specified backup rate, whichever is of higher frequency (Fig. 7-2). ACMV is used commonly for initiation of mechanical ventilation because it ensures a backup minute ventilation in the absence of an intact respiratory drive and allows for synchronization of the ventilator cycle with the patient's inspiratory effort.

Problems can arise when ACMV is used in patients with tachypnea resulting from nonrespiratory or nonmetabolic factors such as anxiety, pain, or airway irritation. Respiratory alkalemia may develop and trigger myoclonus or seizures. Dynamic hyperinflation may occur if the patient's respiratory mechanics are such that inadequate time is available for complete exhalation between inspiratory cycles. This can limit venous return, decrease cardiac output, and increase airway pressures, predisposing to barotrauma. ACMV is not effective for weaning patients from mechanical ventilation because it provides full ventilator assistance on each patient-initiated breath.

PCV can be used to provide ventilator support either with ACMV triggering (PCV-ACM) or SIMV triggering (PCV-SIMV). In contrast to conventional ACMV or SIMV, which are volume-cycled and pressure-limited, PCV-ACM and PCV-SIMV are time-cycled and pressure-limited. During the inspiratory phase, a given pressure is imposed at the airway opening, and the pressure remains at this user-specified level throughout inspiration (Fig. 7-3). Since inspiratory airway pressure is specified by the operator, tidal volume and inspiratory flow rate are dependent rather than independent variables and are not user-specified. PCV is used commonly for patients with documented barotrauma because airway pressures can be limited, as well as for postoperative thoracic surgical patients, in whom the stress across a fresh suture line can be limited. When PCV is used, minute ventilation and tidal volume must be monitored; minute ventilation is varied by the user through changes in rate or in the pressure-controlled value.

PCV with the use of a prolonged inspiratory time is frequently applied to patients with severe hypoxemic respiratory failure. This approach, called pressure-controlled inverse inspiratory-to-expiratory ratio ventilation (PCIRV), increases mean distending pressures without increasing peak airway pressures. It is thought to work in conjunction with PEEP to open collapsed alveoli and improve oxygenation. In acute lung injury (ALI), PCIRV may be associated with fewer deleterious effects than conventional volume-cycled ventilation, which requires higher peak airway pressures to achieve an equivalent reduction in shunt fraction, but there are no convincing data to show that PCIRV improves outcomes in ALI or adult respiratory distress syndrome.^8,9

PSV is a form of ventilation that is patient-triggered, flow-cycled, and pressure-limited; it is designed specifically for use in the weaning process but is also used commonly to ventilate patients with new-onset respiratory failure who are agitated and become asynchronous with other modes of ventilator support. During PSV, the inspiratory phase is terminated when the inspiratory flow rate falls below a certain level; in most ventilators, this flow rate cannot be adjusted by the operator. When PSV is used, patients receive ventilator assist only when the ventilator detects an inspiratory effort (Fig. 7-4). Thus it is mandatory that the patient have an intact respiratory drive to be maintained safely on this mode of ventilation without additional backup ventilator support.

PSV is well tolerated by most patients who are being weaned. PSV parameters can be set to provide full or nearly full ventilator support that can be withdrawn slowly over a period of days in a systematic fashion to gradually load the respiratory muscles and allow the patient to fully resume spontaneous breathing.

CPAP is not a true support mode of ventilation because ventilation occurs through the patient's spontaneous efforts. The ventilator provides fresh gas to the breathing circuit with each inspiration and charges the circuit to a constant operator-specified pressure that can range from 0 to 20 cm H₂O (Fig. 7-5). CPAP is used to assess extubation potential in patients who have been weaned effectively and are requiring little ventilator support.

One of the most serious complications associated with mechanical ventilation is ventilator-associated pneumonia (VAP). VAP is defined as a pulmonary parenchymal infection that has developed after at least 48 hours of mechanical support.¹⁰ It is the most common infection acquired in the ICU, accounting for up to 50% of the antibiotics prescribed in this setting, and has an attributable mortality rate between 5.8% and 27%.^11-15 Because of these ominous statistics, there is great interest in disease pathogenesis and prevention.

Several mechanisms have been proposed for the development of VAP. The most common is pathogenic colonization of the aerodigestive tract. From this point, bacteria can enter the lung via the blood and lymphatics or may be introduced into the airways through or around the endotracheal tube (ETT).^10,16–18 Based upon these proposed mechanisms, several preventive strategies have been implemented to reduce infection. These include decontamination of the aerodigestive tract via topical/oral antimicrobial medications, elevation of the head of the bed, sedation holidays, and aggressively implemented weaning protocols.¹⁹²⁵ The latter two are focused not on interrupting the translocation of bacteria but rather on shortening the duration of mechanical ventilation during which a patient is at risk for VAP.

Several studies have now demonstrated that individual components of this VAP prevention strategy may reduce the incidence of ventilator-associated infection. These observations have led to the creation of VAP prevention bundles with the goal of leveraging the complementary nature of these interventions.²⁶ This bundle includes:

1. Head of bed elevation

2. Oral decontamination

3. Sedation holidays

4. Protocolized ventilator weaning

Recently, Morris and colleagues reported that following the implementation of such a protocol, the rate of VAP decreased from 32 cases per 1000 ventilator days to 12 cases per 1000 ventilator days (p < 0.001).²⁶ These and other similar investigations strongly support the use of VAP prevention bundles as a routine part of clinical care in the ICU.

As mentioned, one of the proposed mechanisms of bacterial entry into the lung is by the endotracheal tube, more specifically around the endotracheal tube. By design, the endotracheal tube has a low-pressure high-volume inflatable cuff on its distal end that provides tube stability and reduces the leakage of air. Although this cuff facilitates ventilatory strategies such as increased PEEP or driving pressures, it also serves as a local nidus of secretions that have collected above it in the proximal trachea. In attempt to mitigate the potential for the drainage of secretions into the lung from this source, manufacturers offer endotracheal tubes that have continuous suction ports in this space to evacuate pooled secretions. Such devices have demonstrated efficacy in reducing VAP and are increasingly being adopted in the ICU.

Finally, there is increasing evidence that some prophylactic interventions commonly used in the inpatient setting may have unintended consequences. Recently, Herzig et al.²⁷ demonstrated that in a single-center survey of over 63,000 hospital admissions, medical suppression of gastric acid was associated with a 30% increased odds of hospital-acquired pneumonia. These data and other investigations linking pharmacologic acid suppression to VAP suggest that altering the pH of gastric secretions may facilitate pathologic colonization.^28,29 Further investigation is required to determined how best to weigh the apparent risks of these medications with their obvious benefits.

A commonly used metric to assess the severity of hypoxemic respiratory failure is the ratio of the partial pressure of oxygen in the arterial blood (Pao₂) divided by the fraction of inspired oxygen (Fio₂: ranging from 0.21 or ambient gas to 1 which corresponds to 100% oxygen). A patient with a Pao₂/Fio₂ ratio less than 300 but more than 200 is categorized as having ALI whereas those with a ratio less than or equal to 200 is categorized as having acute respiratory distress syndrome (ARDS).³⁰ It is estimated that the mortality associated with ARDS is between 25% and 40%.^31,32 And while it is increasingly clear that one of the greatest risks for the development of this condition is the presence of ALI, there are no specific recommendations pertaining to ventilatory strategies in this cohort.

ARDS is characterized by an admixture of shunt and physiologic dead space due to interstitial edema and alveolar exudate caused by dysfunction in pulmonary surfactant. In aggregate, this results in a lung with reduced compliance with regions of profound derecruitment at normal ventilatory pressures. There have been several investigations into the care of such patients. One of the largest was funded by the National Heart, Lung, and Blood Institute and performed by the Acute Respiratory Distress Syndrome Network (ARDSNet).³² A common feature of many of these studies is that the best approach for mechanical ventilation, one that reduces morbidity and mortality, is one that yields an “open” lung at end exhalation that is not subsequently over distended at full inflation.

The open lung strategy of ventilation is thought to work by recruiting damaged, collapsed alveoli and improving oxygenation without causing overdistention and ventilator-associated lung injury in less damaged areas of lung.³³ Open lung ventilation is not a distinct mode of ventilation. Rather, it is a strategy for applying either volume-cycled or pressure-controlled ventilation to patients with severe respiratory failure. The level of PEEP is selected to help avoid cyclic opening and closing of alveolar units, thereby allowing the majority of alveoli to remain inflated during tidal ventilation. Achievement of eucapnia and normal blood pH through adjustments in ventilator tidal volume and breathing frequency are of lower priority. Hypercapnia and consequent respiratory acidosis tend to be well tolerated physiologically, except in patients with significant hemodynamic compromise, ventricular dysfunction, cardiac dysrhythmias, or increased intracranial pressure. Furthermore, data suggest that hypercapnia may have direct beneficial anti-inflammatory effects in the inflamed lung.³⁴ Open lung ventilation has been shown to primarily benefit medical ICU patients with hypoxemic respiratory failure owing to ALI. Although open lung ventilation has not been specifically evaluated in the thoracic surgery patient population, it is reasonable to presume that similar benefits will be observed in thoracic surgery patients with this condition.

Other nonconventional strategies that have been used to manage critically ill patients with hypoxemic respiratory failure include prone-position ventilation, nitric oxide supplementation, inhaled prostacyclin, and extracorporeal membrane oxygenation.³⁵ Each approach has been shown to produce physiologic benefit in uncontrolled observational studies. Results from randomized trials have been less convincing. Nevertheless, a trial of one or more of these approaches may be appropriate in the postoperative thoracic surgery patient with refractory hypoxemia.

Prone positioning is an alternative therapy for patients with refractory hypoxemia. It is effective, simple to implement, and while it has not been shown to reduce mortality, it often produces dramatic and immediate improvements in oxygenation.³⁶ In contrast to its use in the medical ICU patient, prone ventilation will not be appropriate for many postoperative thoracic surgery patients. It may not be safe or practical to prone patients who have recently undergone pneumonectomies or chest wall resections or have multiple anterior chest tubes. Nevertheless, in patients in whom proning is safe, a trial should be considered for the severely hypoxic patient. Prone positioning can be used in combination with any mode of ventilator support. Patients are rotated from the supine position to the prone swimmer's position. Pads are required for bony dependent areas such as the forehead, elbows, ankles, and knees, and frequent turning is required to prevent skin breakdown.

The physiologic basis for improvement in gas exchange after proning appears to result from improved V̇/Q̇ matching in newly dependent lung zones and recruitment of damaged areas of lung through gravity-assisted redistribution of transpulmonary pressures.²² Several small nonrandomized trials have demonstrated improvements in oxygenation with proning. Results of a recently completed randomized clinical trial confirmed these observations but failed to demonstrate a mortality benefit from an initial 10-day course of prone ventilation.³⁷ Nevertheless, prone ventilation still may improve oxygenation in selected postoperative patients for whom it can be instituted safely and thus should be considered as a supplement to conventional ventilation for the patient with refractory hypoxemia.

Inhaled nitric oxide (NO: 2–40 ppm) also has been combined with mechanical ventilation to treat patients with hypoxemic respiratory failure.³⁸ Benefits in oxygenation and reductions in pulmonary artery pressures have been observed in many patients receiving NO, although physiologic benefit has not been universal, and clinical parameters to identify responders have not yet been identified.³⁹ Two randomized clinical trials conducted in patients with hypoxemic respiratory failure have failed to show mortality benefits or decreased resource utilization from NO therapy.⁴⁰ Although not universally effective, inhaled NO still may be effective in selected patients with postoperative hypoxemic respiratory failure. The primary limitation to its widespread implementation is cost which has led many centers to embrace other inhaled therapies for refractory hypoxemia such as prostacyclin.

Inhaled prostacyclin (PGI₂: 0–50 ng/kg/min) is a pulmonary vasodilator that promotes smooth muscle relaxation by increasing cytoplasmic adenosine monophosphate. Its half-life is less than 5 minutes and it has effects similar to those observed with inhaled NO.³⁵ Prostacyclin is formulated to be administered intravenously or via inhalation but the former limits its beneficial impact on ventilation perfusion matching and has the disadvantage of causing systemic hypotension. Although there have been no large randomized trials of inhaled prostacyclin, it is simple to use, requires no special monitoring for administration, and is less expensive than NO.

Extracorporeal membrane oxygenation also may be used in the management of the posttransplantation patient with severe hypoxemia secondary to ischemia-reperfusion injury or severe acute rejection. Although randomized trials have failed to demonstrate consistent benefits from use of extracorporeal membrane oxygenation for established ARDS, posttransplantation hypoxemic respiratory failure has been managed successfully using this approach.^41,42

Mechanical ventilation can successfully support ventilation and gas exchange during surgery, the perioperative period, and during episodes of acute respiratory failure, but it is associated with considerable risks and complications that increase with the duration of intubation.^7,43 Therefore, it is imperative to recognize when sufficient improvement occurs to expeditiously shift attention to liberation from mechanical ventilation. The clinician must carefully weigh the benefits of rapid liberation against the risks of premature trials of spontaneous breathing and extubation. For the majority of patients (∼75%), there is no need for a step-by-step prolonged decrease in ventilatory support. Several studies have demonstrated that the process can be carried out safely by performing three simple steps⁴³⁴⁵: (1) Establish readiness for spontaneous breathing trial (SBT) by recognizing adequate recovery of gas exchange and respiratory muscle function followed by; (2) SBT; and (3) extubation (Fig. 7-6).

The clinician should consider each step as a test with a minimal threshold that has to be met before proceeding with weaning. The assessment for readiness is conducted on a daily²² (or twice daily) basis as soon as mechanical ventilation is initiated. The critical step in this process is to recognize when the causes(s) of respiratory failure have reversed (Table 7-2) to a degree that makes spontaneous breathing possible.

Up to 20% of mechanically ventilated patients may have several failed SBT attempts, despite efforts to reverse the initial cause(s) of respiratory failure (Table 7-2). In these cases it is reasonable to consider tracheostomy to transition to a more gradual withdrawal of ventilatory support. This should be considered earlier when upper airway obstruction, coma, or irreversible or slowly reversible neuromuscular disease complicates clinical care. Tracheostomy tubes have several benefits including patient comfort which affords a reduction in sedation, better oral care, possible speech restoration, provision of more secure airway, and reductions in the work of breathing.

The guidelines outlined in the preceding section are satisfactory for weaning most thoracic surgery patients from mechanical ventilation. Patients who have undergone a pneumonectomy, however, often present special challenges that can slow weaning and complicate postoperative management. Shift of the mediastinum toward the remaining lung owing to rapid accumulation of fluid within the hemithorax can increase pleural pressures and thus decrease transpulmonary distending pressures in the remaining lung. This imposes an added, severe restrictive physiologic load on the respiratory system postoperatively that can precipitate respiratory failure. Destabilization of the mediastinum can have additional adverse consequences on lung physiology. The loss of transpulmonary distending pressures owing to mediastinal shift can lead to collapse of small airways and can produce wheezing on examination. This pseudo-obstruction pattern tends to be refractory to bronchodilators and compromises breathing by increasing the work of breathing. Therapeutic drainage of fluid from the side of the pneumonectomy, restoration of pressure equilibrium between the hemithoraces, and reestablishment of a midline mediastinal position lead to resolution of this problem.

Hemodynamic instability also may occur after pneumonectomy because of several factors that influence right-sided heart function. Loss of a portion of the functioning pulmonary vascular bed may contribute to increased pulmonary pressures in patients with limited preexisting reserve. Lowering the intrathoracic pressure by removing excess air or fluid from the pneumonectomized side can acutely increase right ventricular afterload and contribute to right-sided heart failure. Conversely, increased intrathoracic pressures resulting from rapid accumulation of fluid on the pneumonectomized side can decrease preload and cause hypotension and reduced cardiac output. Reinstituting drainage of the chest cavity to atmospheric pressure via a chest tube connected to a water seal represents effective management.

We wish to acknowledge Dr. Edward P. Ingenito for his contribution to prior versions of this chapter.

This chapter has been specifically geared to the novice, although everyone will find value in its pages. A simple description of the anatomy of the ventilator is a good place to start to understand this complex issue. The authors then explain the conventional modes of ventilation and explain the acronyms that many students find confusing. There is a lot of practical information in this chapter, including positioning of the patient, sedation management, and special considerations for ventilating the pneumonectomy patient. This chapter is written in a clear style that makes it of practical value to anyone working in a thoracic ICU.