| Postoperative Complications | Respiratory Care| Cardiovascular Therapy| Fluid Management| Nutrition in Critical Patients |Traumatic Brain Injury-Stroke-Brain Death |

Specific considerations of respiratory pathology and care in the neurosurgical intensive care unit (NICU) are discussed. The development of pulmonary complications contributes significantly to mortality in critically ill neurosurgical patients and worsens neurologic outcome. Prompt recognition and treatment of pulmonary disorders are therefore important in the management of neurosurgical patients.

I. Respiratory disorders

Clinical physiology of oxygen transfer. Arterial hypoxemia commonly occurs in patients after central nervous system (CNS) injury, adversely affecting neurologic outcome, especially in patients who have sustained head trauma. Arterial hypoxemia (PaO₂ <50 mm Hg) causes cerebral vasodilatation and increases in cerebral blood flow and cerebral blood volume, which could have a detrimental impact on intracranial pressure (ICP). The major causes of arterial hypoxemia are (a) a decrease in alveolar oxygen tension (Pao₂) owing either to hypoventilation or to decreased inspired oxygen tension (Pio₂) or (b) an increase in alveolar-arterial oxygen tension difference (P[A-a]O₂). Inspired gas is mixed with gas in the functional residual capacity (FRC) to make up alveolar gas. Mixed venous blood equilibrates with alveolar gas and is distributed to the systemic circulation as mixed arterial blood.
Causes of decreased Pao₂. Pao₂, the major driving factor for arterial oxygenation, may be reduced because of either a decrease in Pio₂ or diminished alveolar ventilation from respiratory depression, increased dead space, or mechanical impairment. Pao₂ is determined by the barometric pressure (PB), the fraction of inspired oxygen (Fio₂), and the ratio of oxygen consumption ([V with dot above]O₂) to alveolar ventilation ([V with dot above]_A):

PAO₂ = PB x ( FIO₂  - VO₂ /VA).

As far as [V with dot above]A is directly proportional to carbon dioxide production
([V with dot above]CO₂), [V with dot above]A = [V with dot above]CO₂/PACO₂ × (k), and PACO₂ ~ PaCO₂,
Equation .1 can be transformed into the alveolar gas equation .2:

PAO₂ = FIO₂(PB-47)- (PaCO2/R_q).
where R_Q = respiratory quotient, R_Q = [V with dot above]CO₂/[V with dot above]O₂ (0.8 in resting state), and 47 = saturated water vapor pressure at the PB in mm Hg.
Respiratory depression (hypoventilation). Respiratory depression can be either central (from CNS disease, narcotics, or sedatives) or peripheral (from neuromuscular disease, muscle relaxants).
1. Traumatic brain injury (TBI). Variations in both the depth and rate of spontaneous respirations occur in 60% of patients who have CNS injury. Five respiratory patterns are observed in patients who have head injuries and brain tumors (Fig.-1). Pathologic hypoventilation and hyperventilation lead to hypoxemia, pH disturbances, and electrolyte imbalance.

2. Spinal cord injury (SCI). High cervical spinal cord lesions impair ventilation because of either partial or complete loss of innervation of the diaphragm (C3-5), with loss of the ability to cough. The diaphragm is spared when injury is below C6-7. However, injury at the thoracic level results in the loss of intercostal and abdominal muscle function, which reduces FRC and creates paradoxical ventilation (chest retraction on inspiration and expansion during expiration), loss of cough, and reduced ability to handle secretions. All lung volumes, except residual volume, are markedly decreased. With a C5-6 lesion, ventilation may be inadequate and forced vital capacity (FVC) decreased to 30% of predicted capacity. Improvement of FVC occurs with time because of the strengthening of the accessory muscles (clavicular portion of the pectoralis major and sternocleidomastoid), development of spasticity (which stops the paradoxical chest wall motion), improvement of diaphragmatic function, and decreased usage of narcotics and sedatives. At 5 months after injury, FVC reaches 60% of predicted values.
3. Associated airway injuries in multiple trauma patients including tracheobronchial fistula, tracheobronchial disruption, and rupture of diaphragm cause acute hypoventilation and profound hypoxemia.
Increased airway dead space. Dead space increases with drug administration (such as atropine and nitroprusside for deliberate hypotension), mechanical ventilation with positive end-expiratory pressure (PEEP), hypovolemia, hypotension, hypocapnia, and cardiopulmonary disease.
Mechanical impairment. Airway obstruction from oropharyngeal soft tissue, uncleared secretions, hemoptysis, foreign bodies, bleeding, or bronchospasm can cause mechanical impairment of effective alveolar ventilation. The restriction of free chest movement from surgical positioning, surgical retraction of the chest and abdomen, traumatic rupture of the diaphragm, pneumothorax, hemothorax, flail chest and restrictive lung disease can reduce alveolar ventilation significantly. Proper recognition and treatment of traumatic chest injuries are crucial to the successful management of patients after traumatic injury.
Causes of increased P(A-a)O₂. The P(A-a)O₂ may be increased as the result of both pulmonary and extrapulmonary factors. Pulmonary disorders can increase P(A-a)O₂ because of a (a) shunt, (b) [V with dot above]A/[Q with dot above] mismatch, and (c) limitation of diffusion. Most pulmonary disorders increase the [V with dot above]_A/[Q with dot above] mismatch and shunt but seldom affect diffusion capacity. Decreases in cardiac output and either a respiratory or metabolic alkalosis are important extrapulmonary factors that increase P(A-a)O₂ in the presence of lung disease.
Extrapulmonary factors. Two extrapulmonary factors that commonly contribute to hypoxemia in neurosurgical patients who have preexisting lung disease are (a) hyperventilation to reduce ICP and (b) excessive depletion of intravascular volume from fluid restriction and the administration of diuretics.
1. Respiratory alkalosis. Hyperventilation to induce hypocapnia results in a decrease in venous return and cardiac output. Hypocapnia increases pulmonary shunt, especially in the presence of lung disease. The mechanism is unclear but might be related to hypocapnic bronchoconstriction and inhibition of hypoxic pulmonary vasoconstriction. In addition, the oxyhemoglobin dissociation curve shifts to the left, resulting in a lower Pa O₂ for any given oxygen saturation (Fig.-2). This is especially noticeable when the Pa O2 is equal to or <70 mm Hg.

Figure -2. Shift in the oxyhemoglobin dissociation curve with change in blood pH.

2. Hypovolemia and decrease in P[ v with bar above]O₂. Pa O₂ is lower when the mixed venous oxygen tension (P[v with bar above]O₂) is decreased, as can occur with decreased cardiac output, increased oxygen consumption, or severe anemia with increased oxygen extraction. The effect of P[v with bar above]O₂ can be quite significant, especially when there are areas of low [V with dot above]A/[Q with dot above] and shunt (Fig.-3). For instance, in the presence of [V with dot above]A/[Q with dot above] mismatch, PaO₂ = 80 mm Hg when P[v with bar above]O₂ is 50 mm Hg. If P[v with bar above]O₂ is decreased to 20 mm Hg from a decrease in cardiac output, then Pa O₂ decreases to 40 mm Hg without any change in the degree of [V with dot above]A/[Q with dot above] inequality or shunt. Therefore, low cardiac output from hypovolemia or myocardial depression increases P(A-a)O₂ and might result in hypoxemia, especially in the presence of pulmonary disease.
Pulmonary factors that increase P(A-a)O₂ in critically ill neurosurgical patients consist of neurogenic pulmonary edema (NPE), [V with dot above]_A/[Q with dot above] mismatch, and abnormalities of lung parenchyma.

Figure-3. Influence of mixed venous oxygen tension on arterial oxygen tension.

a. Neurogenic etiologies of increased P(A-a)O₂
(1) Reduced FRC. The FRC is often reduced after TBI and SCI at the cervical and thoracic levels. Head trauma patients might exhibit a reduction in FRC and pulmonary compliance and an increase in pulmonary shunting without evidence of pulmonary disease on the chest radiograph.
(2) NPE. Classic NPE develops in severe CNS injury, such as TBI, SCI, and subarachnoid hemorrhage (SAH). Of immediate onset, NPE becomes clinically recognizable 2 to 12 hours after injury. NPE is generally of short duration and resolves within hours to days. The pathogenesis of NPE is complex and related to circulatory hyperactivity owing to catecholamine release, particularly norepinephrine, which causes an increase in the permeability of the alveolar capillary barrier.
(3) Neurogenic alterations in [v with bar above]A/[Q with dot above] matching. Hypoxemia and increased pulmonary shunting have been observed in animals and patients who have CNS disorders in the absence of pulmonary disease. The etiology of these changes in [V with dot above]A/[Q with dot above] regulatory mechanisms is multifactorial. Release of inflammatory and anti-inflammatory mediators in patients after TBI may cause acute lung injury (ALI). Brain thromboplastin is released into the blood with TBI, causing deposition of fibrin microemboli and platelet aggregates in the pulmonary capillaries. Neutrophil accumulation and the release of vasoactive substances might also cause local [V with dot above]_A/[Q with dot above] abnormalities. Patients who have an increase in fibrin spit products are also more likely to develop respiratory failure.
Pulmonary etiologies of increased P(A-a)O₂
(1) Gastric aspiration. Aspiration of gastric contents occurs frequently in comatose neurosurgical patients and in trauma patients because of intoxication from alcohol and street drugs and abdominal injury. Aspiration of large volumes of fluid with a low pH or particulate matter results in aspiration pneumonitis and ALI. The aspiration of large foreign bodies, such as gravel, gum, or teeth, might also cause bronchial obstruction, atelectasis, and hypoxemia, and potentially may lead to necrotizing pneumonia. Treatment of aspiration is primarily supportive. Antibiotics are not indicated.
(2) Atelectasis and lung consolidation are common in the neurosurgical patient. Central respiratory depression, cervical and thoracic spine injury, hypoventilation from pain, altered levels of consciousness, bronchial obstruction from mucous plugs or aspiration of particulate matter, and compression of adjacent lung by traumatic pneumothorax, hemothorax, or pleural effusion all contribute to the formation of atelectasis. Prevention and treatment of lung collapse through hyperventilation, recruitment maneuvers (PEEP, inverse-ratio ventilation where inspiration is longer than expiration, "sigh" breaths), and aggressive physiotherapy can prevent the development of lung infection. Because SCI patients are not able to clear secretions adequately, therapeutic bronchoscopy, bed rotation, and turning the patient into the prone position can help clear secretions.
(3) Nosocomial pneumonia (NP)/ventilator-associated pneumonia (VAP). VAP is an NP that develops in patients 48 hours or more after endotracheal intubation. NP is the second most common (after urinary tract infection) and the most serious infection in hospitalized patients. VAP occurs in 5% to 30% of intensive care unit (ICU) patients and is responsible for 30% to 40% of all ICU-acquired infections. NP/VAP develops in about half of critically ill neurosurgical patients who have isolated TBI and SCI and in nearly two-thirds of patients after cervical SCI. Pulmonary complications, particularly pneumonia, represent an independent risk factor for death in SCI.
(a) Risk factors for NP/VAP are presented in Table-1.
(b) Early onset NP/VAP and late-onset VAP are distinguished by their respective times of onset and pathogens. In the general ICU population, early onset NP/VAP develops within the first 5 days of hospitalization while late-onset VAP develops after 5 days. Early onset NP/VAP is characterized by community-acquired microorganisms including gram-positive cocci (Staphylococcus aureus, Haemophilus influenzae, Streptococcus pneumoniae), Enterobacteriaceae, and anaerobes. In late-onset VAP, the causative pathogens are hospital-acquired (nosocomial), highly virulent, antibiotic-resistant microorganisms (e.g., Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella, methicillin-resistant Staphylococcus aureus [MRSA], vancomycin-resistant Enterococcus). The extreme virulence and the antibiotic resistance of the flora are responsible for the high mortality rate of 40% to 60%, which can go as high as 80% in late-onset VAPs. Community-acquired organisms predominate as the causative agents of pneumonia in critically ill TBI patients in the first 10 days after admission. With comatose patients, VAP develops most frequently between hospital days 4 and 7.

(c) Pathogenesis of late-onset NP/VAP in terms of sources of infection includes major aspirations, nasal colonization by S. aureus, and microaspiration of colonized oral, upper airway, and gastric secretions, all of which can lead to the development of microabscesses in the lung. Hypoventilation with resultant insufficient alveolar aeration and consequent alveolar collapse and consolidation hastens the onset of infection.
(d) Diagnosis (Table-2). Unfortunately, there is no "gold standard" for diagnosis of VAP. Along with clinical criteria, early bacteriological detection of causative microorganisms is important for appropriate antibiotic treatment. If not supported by bacteriological results, empiric antibiotic treatment is often inadequate. The use of invasive bronchoscopic techniques (bronchoalveolar lavage and protective specimen brushing) with quantitative cultures of the specimen allows precise adjustment of antibiotic therapy. However, finalizing quantitative cultures takes 2 to 3 days, which inevitably leads to delay in instituting appropriate treatment.
 

(e) Choice of antibiotic treatment depends on onset of NP/VAP, risk factors, and the severity of the illness. In general, in early onset NP/VAP, monotherapy with amoxyclavulanate, a second-generation or nonpseudomonal third-generation cephalosporin, or carbapenem for 5 to 7 days is effective in early onset NP/VAP. Late VAP requires broad coverage of resistant pathogens, and empiric combination therapy with two or three antibiotics should be initiated with an antipseudomonal cephalosporin or carbapenem with amikacin or quinolone. For MRSA coverage, linezolid or vancomycin should be added, although linezolid is more effective than vancomycin. After quantitative cultures have been finalized, empiric antibiotic coverage should be adjusted to the particular pathogen and continued for at least 7 to 21 days, depending on the microorganism and the severity of illness.
(f) Preventive strategies for VAP are directed at preventing aerodigestive tract colonization, aspiration, and lung consolidation. Routine hand washing with either water and soap or alcohol gel is the most effective strategy for reducing colonization of nosocomial infections. Other modalities include restriction of antibiotic usage, attention to oral hygiene, isolation of patients who have antibiotic-resistant pathogens, and selective digestive decontamination. Semirecumbent positioning, subglottic drainage, prevention of unexpected extubation, and avoidance of nasal intubation, gastric overdistention, and oversedation can reduce the chance of aspiration.
(4) Acute lung injury (ALI)/Adult respiratory distress syndrome (ARDS)
(a) Definition. ALI/ARDS is a clinical syndrome of diffuse lung injury that results in noncardiogenic, high-permeability pulmonary edema. ALI/ARDS consists of the following: (a) acute onset, (b) bilateral pulmonary infiltrates on the chest radiograph, (c) noncardiogenic origin of edema: pulmonary capillary occlusion pressure of <18 mm Hg, (d) hypoxemia, as defined by a ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen (Pa O₂/FIO₂) that is ≤300 in ALI, and ≤200 in ARDS. The incidence of ALI/ARDS in North America and Europe varies from 1.5 to 75 cases/100,000 population. ALI/ARDS is a severe disease with a significant mortality rate of 22% to 32% for ALI and approximately 40% to 70% for ARDS. One-third of ALI cases progress to ARDS in the first week, particularly in trauma patients. One-third of patients who have isolated TBI develop ALI/ARDS; this significantly increases mortality and worsens neurologic outcome. The major cause of death in patients with ALI/ARDS is sepsis and multiorgan failure. Only approximately 10% to 20% of patients dying from ARDS die from hypoxemia per se.
(b) Risk factors (Table-3). Primary ARDS is caused by direct lung injury and primarily pneumonia and has a higher mortality rate than that of secondary ARDS, which is caused by indirect lung injury.
(c) Pathogenesis and clinical course. ALI/ARDS usually develops within the first 72 hours of the initiating event with 60% to 70% of cases occurring within the first 24 hours. The initial step in the pathogenesis of ALI/ARDS is endothelial cell activation in the pulmonary microvasculature by various insulting agents such as microbial products, different cytokines, and histamine, which leads to microcirculatory distress, microcoagulation, neutrophil activation and imbalance between proinflammatory and anti-inflammatory mediators, impairment of the alveolar-capillary barrier, and influx of protein-rich fluid into the alveoli. The alveolar epithelium is damaged, resulting in the disruption of the epithelial ion and fluid transport system, impairment of surfactant production, and loss of the epithelial barrier. Disorganized epithelial repair leads to fibrosis. The degree of epithelial injury correlates with the severity of illness and outcome. Significant abnormalities of pulmonary vascular permeability usually persist for 1 week after the onset of ARDS. The acute phase may either resolve in the first week or progress to fibrosis with persistent hypoxemia. Patients who do not improve in the first week have a poor prognosis. In late phases, pulmonary hypertension can develop from elevated pulmonary vascular resistance, primarily from arterio-capillary obstruction, rather than pulmonary vasoconstriction. Pulmonary compliance decreases. The chest computed tomographic (CT) scan often shows diffuse interstitial thickening and honeycombing. In survivors of ARDS, muscle wasting and weakness are the most prominent extrapulmonary conditions that persist for a year after recovery.

Ventilator-induced lung injury (VILI)/ventilator-associated lung injury is an inflammatory syndrome of diffuse alveolar damage that worsens the outcome in patients who have ALI/ARDS. Several mechanisms are suggested for the pathophysiology of VILI: oxygen toxicity, alveolar overdistension from either high tidal volumes (volutrauma) or high airway pressures (barotrauma), and repeated alveolar opening and collapse during the ventilatory cycle. Because ALI/ARDS is a heterogenous lung disease, positive-pressure ventilation causes uneven distribution of volume and alveolar pressure to the extent that even moderate tidal volumes can overstretch healthy alveoli, leading to inflammatory responses and exacerbation of inflammatory processes. PEEP diminishes VILI by recruiting alveoli and preventing alveolar collapse.
Treatment. Lung protective strategies for mechanical ventilation have contributed to the decline in mortality along with the prevention and treatment of nosocomial infection such as pneumonia and sepsis, prophylaxis against upper gastrointestinal bleeding, and prevention of thromboembolism.
Lung-protective ventilation is the cornerstone of ALI/ARDS management. Positive-pressure ventilation with a tidal volume of 6 mL/kg of ideal body weight and plateau pressure <30 cm H2O significantly decreased mortality compared to conventional ventilation. PEEP should be adjusted to allow for an Fio₂ of <0.6 to prevent oxygen toxicity. The oxygenation goal is either a Pa O₂ of >60 mm Hg or an oxygen saturation of 90% or better.
Permissive hypercarbia. Lung-protective ventilation can inevitably lead to hypoventilation. Hypercarbia with a Paco₂ of 60 mm Hg to 100 mm Hg and a pH of 7.25 can be safely tolerated. However, because CO₂ is a potent cerebral vasodilator, permissive hypercarbia is contraindicated in neurosurgical patients who have intracranial hypertension.
Prone position. Turning the patient from the supine to the prone position improves oxygenation through favorable redistribution of ventilation-perfusion matching by recruitment of collapsed alveoli. In patients who have SAH and normal ICP, the prone position enhances brain tissue oxygenation but can increase ICP slightly. Risks of the prone position include accidental extubation, loss of intravenous access, hemodynamic instability, coughing, and late complications such as edema and pressure sores.
Nitric oxide (NO). Although not proven to affect survival, inhaled NO improves oxygenation and can be used in life-threatening hypoxemia unresponsive to other maneuvers. NO can be used safely in patients who have increased ICP and in situations where it is not possible to turn the patient to the prone position.
Diuresis and fluid restriction. Regardless of origin, diuresis and fluid restriction are beneficial in ALI/ARDS but may be impossible to achieve in patients who have vasospasm after SAH and require hypertensive hypervolemic hemodilution (triple-H) therapy.
Glucocorticosteroids. Moderate doses of glucocorticosteroids in the late phase of hypoxemic ARDS may suppress fibroproliferation in the lung and could be beneficial in patients who do not have signs of infection.
Cardiogenic pulmonary edema. Pulmonary edema of cardiogenic etiology can occur in the neurosurgical patient. Left ventricular dysfunction may be from myocardial ischemia or infarction, adverse drug reactions, tachyarrhythmias, severe hypertension, or fluid overload. Older patients are at greater risk; however, myocardial ischemia from severe anemia, hypotension, fluid overload, excessive mannitol administration, and triple-H therapy can precipitate pulmonary edema in the younger patient. Treatment involves diuresis, fluid restriction, and supportive care.
Pulmonary embolism. Pulmonary emboli are common in critically ill neurosurgical patients. In SCI, the incidence of deep vein thrombosis reaches 100% without prophylaxis and 10% to 25% with prophylaxis. Pulmonary emboli occur in >10% of SCI patients and usually originate from venous thrombosis in lower extremities and, less commonly, from either the pelvic or prostatic venous plexi or the right ventricle. Pulmonary embolism causes acute pulmonary hypertension, right ventricular failure, and a reduction in cardiac output. The diagnosis of smaller pulmonary emboli should be considered in patients who have unexplained tachypnea and chest pain. The Pa O₂ is usually normal, and the Paco2 is decreased in the tachypneic patient who has small pulmonary emboli. Massive pulmonary embolism causes acute cardiorespiratory failure. In the past decade, the multidetector row spiral CT scan has become the first-line diagnostic tool for the diagnosis of pulmonary embolism. If the multidetector row spiral CT scan is not available, CT scan and pulmonary angiography are indicated. Definitive treatment is predominantly supportive and involves anticoagulation. If anticoagulation is contraindicated, an inferior vena cava filter should be placed.

II. Respiratory care

Administration of oxygen and airway management. All TBI patients should be considered hypoxic until proven otherwise, and supplemental oxygen (O₂) should be administered. A Pa O₂ of >60 mm Hg is necessary to maintain adequate cerebral oxygenation in patients after TBI and SCI.
Endotracheal intubation is indicated in the neurosurgical patient for the following conditions:
Inability to protect the airway or clear secretions
Need to reduce ICP by control of ventilation
PaO₂ <60 mm Hg in spite of supplemental O₂ by mask
Paco₂ >50 mm Hg
pH <7.2
Respiratory rate >40/minute or <10/minute
Tidal volume <3.5 mL/kg
Vital capacity <10 to 15 mL/kg
Muscle fatigue
Airway compromise
Hemodynamic instability
Orotracheal intubation. Orotracheal intubation is preferable for urgent intubation in the presence of increased ICP, hypoxemia, and/or hemodynamic instability. In the hemodynamically stable patient, either thiopental, 3 to 5 mg/kg, or propofol, 2.5 mg/kg, and succinylcholine, 1.5 mg/kg, can be used to prevent coughing and an increase in ICP. Alternatively, rocuronium, 1 to 1.5 mg/kg, can be used for muscle paralysis. Etomidate, 0.1 to 0.3 mg/kg and lidocaine, 1 to 1.5 mg/kg, can be used in the less stable patient. With patients who have a full stomach, a modified rapid sequence induction with cricoid pressure and ventilation by mask might be used to avoid marked increases in ICP from hypoxemia and hypercarbia. The head and neck should be stabilized by an assistant (manual in-line stabilization) to prevent head movement in the presence of potential cervical spine injury. Manual in-line stabilization may make a difficult intubation more difficult and result in more movement of unstable structures lower in the neck.
Nasotracheal intubation. Nasotracheal intubation is of minimal use in NICU patients because of the possibility of resultant obstruction of sinus drainage, which increases the risk of sinusitis, pneumonia, and subsequent sepsis and ALI/ARDS. Nasotracheal intubation is absolutely contraindicated in the presence of a basilar skull fracture. Awake nasotracheal intubation might increase ICP in the head-injured patient.
Tracheotomy is absolutely indicated in the presence of high SCI (higher than C5). Emergency tracheotomy or cricothyrotomy could be required in patients who have TBI and multiple facial fractures rendering bag-and-mask ventilation difficult. Compared with endotracheal intubation, tracheotomy decreases anatomical dead space and the work of breathing slightly, improves access for suctioning, provides more comfort to the patient, and reduces sedation requirements, all of which may be beneficial to patients who have difficulties in weaning from the ventilator. Early tracheotomy is considered in the first week of admission. Although there is no consensus as to the beneficial effects of early surgical tracheotomy, performing a bedside percutaneous dilatational tracheotomy (PDT) on the second or third day after injury significantly reduces the incidence of pneumonia, the ICU length of stay, and the mortality in patients who were expected to need mechanical ventilation for >14 days. Therefore, early bedside PDT should be considered in patients who have significant neurologic injury, absent gag reflex, or inability to protect their airway and clear secretions and for patients who are in a persistent vegetative state. Because of the advantages of bedside PDT, including negation of the need to involve the operating room and to transport a critically ill patient, this technique has gained popularity in recent decades. The procedure requires the participation of two physicians (surgeon, otolaryngologist, intensivist) and consists of cricothyroidotomy and tracheal dilatation using the Seldinger technique with bronchoscopic guidance through the endotracheal tube by an experienced bronchoscopist. Although complications are rare with an experienced team, the major complication is bleeding (1% to 2%).
Invasive mechanical ventilation is used to improve gas exchange after nasal or oral endotracheal intubation. Vital signs should be monitored and arterial blood gases measured 30 minutes after changing ventilator settings.
Controlled mechanical ventilation is used in the operating room and occasionally in the NICU in selected patients who have no ventilatory efforts owing to neuromuscular blockade, high cervical SCI, absent ventilatory drive, drug overdose, or brain death.
For patients in whom spontaneous ventilation is preserved, the preferred assisted modes of ventilation are those in which the ventilator delivers a breath triggered by the patient's inspiratory effort (Fig.-4). Along with classic volume-controlled and pressure-controlled modes of mechanical ventilation (assist-controlled, synchronized and nonsynchronized intermittent mandatory ventilation, and pressure support ventilation), modern ventilators provide a variety of new ventilation modes. These modes include "dual-controlled"� mechanical ventilation with preset tidal volume and inspiratory pressure ("volume-assured pressure support"� or "pressure augmentation"� mode), airway-pressure release ventilation (the mode of pressure-controlled inverse-ratio ventilation, allowing spontaneous breathing), and proportional assist ventilation.
PEEP is used to increase arterial oxygenation by alveolar recruitment and increasing FRC, which decreases intrapulmonary shunting and [V with dot above]_A/[Q with dot above] abnormalities. The application of PEEP decreases the need to give high, potentially toxic, concentrations of oxygen. On the other hand, PEEP increases Paco₂ by increasing dead space ventilation. Hemodynamically, PEEP decreases venous return, cardiac output, and systemic arterial blood pressure. These effects can be partially attenuated by hydration. The varied clinical responses of TBI patients to PEEP might be secondary to PEEP's effect on other hemodynamic and respiratory variables. The use of 10 cm H₂O of PEEP improves oxygenation but can decrease cerebral perfusion from the decrease in cardiac output. PEEP usually causes clinically inconsequential increases in ICP in patients who have severe TBI. The influence of PEEP on ICP is less prominent in patients who have stiff lungs (as with ALI/ARDS) and are presumably the ones who need PEEP the most. A PEEP of up to 30 cm H₂O has been used to treat hypoxemia with minimal adverse effect on ICP. However, because significant potentially serious increases in ICP and clinical deterioration have been observed in some TBI patients after the application of PEEP, ICP should be monitored when PEEP is used. Increases in ICP from PEEP can be reduced by elevating the head 30° and concurrently administering mannitol. The abrupt removal of PEEP might also increase ICP.

Figure-4. Modes of mechanical ventilation.

Paw, airway pressure; SB, spontaneous breathing; PCV, pressure-controlled ventilation; PEEP, positive end-expiratory pressure; VCV, volume-controlled ventilation; VC-IRV, volume controlled-inverse ratio ventilation; I, inspirium; E, expirium; sb, spontaneous breath; T, trigger; CPAP, continuous positive airway pressure; PSV, pressure support ventilation; APRV, airway pressure release ventilation.

Noninvasive mechanical ventilation is contraindicated for neurosurgical patients in acute respiratory failure who do not have an endotracheal tube. This modality is useful only for patients who, in the absence of neurologic deficits, have, for example, an exacerbation of chronic obstructive pulmonary disease.
Monitoring patients receiving mechanical ventilation. The efficacy and safety of mechanical ventilation in patients is evaluated by monitoring gas exchange, airway pressure, breathing pattern, lung compliance, hemodynamic function, chest radiographs, and chest CT scans, the "gold standard"� for diagnosis of lung pathology.
Weaning. Clinically stable patients should be frequently reevaluated for readiness for a trial of weaning. However, standard weaning criteria, including normal neurologic status, are not applicable to head-injured patients, in whom impaired neurologic status can delay weaning and extubation.
Clinical parameters to be met before attempting weaning
1. Cause for instituting mechanical ventilation is sufficiently resolved.
2. Adequate oxygenation: Pa O₂/Fio₂ ≥ 150 to 200 or hemoglobin saturation (SaO₂) of >90% on Fio₂ ≤ 0.4 to 0.5 with PEEP ≤ 8 cm H₂O, pH ≥ 7.25.
3. Hemodynamic stability (low-dose inotropes/vasopressors are allowed).
Weaning techniques. Spontaneous breathing trial for 30 to 120 minutes should be attempted in patients considered candidates for weaning from mechanical ventilation. The trial can be performed with a T-piece, using a low level of continuous positive pressure (5 cm H₂O) or pressure support (5 to 7 cm H₂O).