Chapter 20: Head Injury

Head injury and its associated neurologic dysfunction are common in the United States and other developed countries. The morbidity and mortality associated with this problem are related to acute cerebral edema and other space occupying lesions that increase intracranial pressure (ICP). Head injury is often traumatic in origin. However, similar effects may be seen with surgical (eg, postcraniotomy for tumor resection) and medical (eg, cerebral vascular accident, postresuscitation hypoxia, hepatic failure) problems.

Physiology

Because the skull is rigid, intracranial volume increases result in an increase in ICP. The relationship between intracranial volume and ICP is described by the cerebral compliance curve (Figure 20-1). Although small increases in intracranial volume are tolerated without an increase in ICP, larger increases in volume result in large increases in ICP. This increase in ICP decreases cerebral blood flow, resulting in cerebral hypoxia. With large increases in ICP, the swelling brain herniates through the tentorium, resulting in compression of the brain stem. Much of the management of head injury relates to efforts to control ICP.

Cerebral perfusion pressure (CPP) is defined as the difference between mean arterial pressure (MAP) and ICP:

Normally, ICP is less than 10 mm Hg and MAP is about 90 mm Hg, resulting in a normal CPP of more than 80 mm Hg. The target CPP is 50 to 70 mm Hg. CPP less than 50 mm Hg should be avoided. In patients with acute head injury, the ICP is frequently measured. CPP is decreased by either a decrease in MAP or an increase in ICP. Thus, treatments that decrease MAP (eg, positive pressure ventilation, diuresis, vasodilator therapy) decrease CPP, whereas treatments that decrease ICP (hyperventilation, mannitol) increase CPP. A normal physiologic response to an acute increase in ICP is hypertension with bradycardia, which is called the Cushing response.

Mechanical ventilation can increase ICP and decrease CPP because the increased intrathoracic pressure associated with mechanical ventilation. Positive end-expiratory pressure (PEEP) has the potential of decreasing MAP and venous return. A decrease in venous return increases ICP and a decrease in MAP decreases CPP.

Clinical Findings

Increases in ICP can result in abnormal ventilatory patterns such as Cheyne-Stokes breathing, central neurogenic hyperventilation, and apnea with severe injury. Compression of the brainstem (ie, transtentorial herniation) results in dilated nonreactive pupils, posturing (decerebrate and decorticate), and cardiovascular collapse.

Neurogenic Pulmonary Edema

Acute head injury and elevated ICP can result in neurogenic pulmonary edema (NPE). NPE is a noncardiogenic pulmonary edema and is clinically indistinguishable from acute respiratory distress syndrome (ARDS). It results in a decreased functional residual capacity, decreased lung compliance, increased intrapulmonary shunt, and hypoxemia. Treatment of NPE is similar to that of other causes of ARDS, including oxygen therapy and PEEP.

Management

Management of acute head injury involves both hemodynamic and respiratory management. Techniques to control ICP are briefly summarized in Table 20-1. Hemodynamic control of arterial blood pressure is important to maintain CPP. Respiratory management involves maintenance of an adequate Paco₂ and Pao₂. Care must be taken to avoid a high MAP, which can adversely affect CPP by decreasing venous return (resulting in an increase in ICP) and decreasing cardiac output (resulting in a decrease in MAP).

In the past, respiratory care of the patient with head injury included the use of iatrogenic hyperventilation. However, this therapy has not been shown to increase survival and is no longer recommended. For an acute increase in ICP, the patient may be temporarily hyperventilated until definitive therapy is instituted, after which the Paco₂ is gradually restored to normal. Care must be taken to avoid rapid increases in Paco₂, which may produce dangerous increases in ICP.

Paco₂ has an indirect effect on cerebral vascular tone due to its effect on pH. It is the change in pH that affects the tone of cerebral blood vessels and thus, cerebral blood volume and ICP. A decrease in pH (increased Paco₂) causes cerebral vascular dilatation and an increase in ICP. An increase in pH (decreased Paco₂) causes a decrease in ICP. During hyperventilation therapy, the brain quickly equilibrates to changes in Paco₂ and a new steady state is established within 4 to 6 hours and over time the pH normalizes reducing the benefit of the decreased Paco₂. Although iatrogenic hyperventilation is not recommended, permissive hypercapnia may be associated with unacceptable elevations in ICP.

As shown in Figure 20-2, increases in Paco₂ and decreases in Pao₂ result in increases in ICP. Thus, normal oxygenation and acid-base balance are goals of ventilation in patients with an increased ICP. Increases in alveolar pressure may result in an increase in ICP due to a decrease in venous return and a decrease in cardiac output.

Indications

Indications for mechanical ventilation in patients with head injury are listed in Table 20-2. The most common reason to ventilate these patients is central respiratory depression due to the primary injury. In such patients, lung function may be near normal and mechanical ventilation is straightforward. In patients with traumatic injury, associated injuries to the spine, chest, and abdomen may also require the initiation of mechanical ventilation. Positive pressure ventilation may be necessary due to neurogenic pulmonary edema. Finally, some therapies for acute head injury (eg, barbiturates, sedation, and paralysis) result in central respiratory depression, necessitating mechanical ventilation.

Ventilator Settings

Recommendations for initial ventilator settings for patients with head injury are listed in Table 20-3 and Figure 20-3. Full ventilator support is almost always initially required for these patients, and can be provided by continuous mandatory ventilation (A/C). Because of the depressed neurologic status of these patients and the need to control Paco₂, pressure support ventilation as the initial mode in these patients is usually not appropriate. As respiratory status improves and spontaneous breathing becomes acceptable, pressure support ventilation can be used.

Because patients with head injury often have relatively normal lung function, oxygenation is usually not a problem. With these patients, 100% oxygen is initially administered and can be rapidly weaned using pulse oximetry. A Pao₂ more than 80 mm Hg is often used because this minimizes the potential for periodic episodes of hypoxemia and associated rises in ICP. An initial PEEP level of 5 cm H₂O is usually appropriate and adequate. Although there is concern related to the effects of PEEP on ICP, PEEP usually does not adversely affect ICP at levels less than or equal to 10 cm H₂O. With neurogenic pulmonary edema, the management of oxygenation is similar to that with other causes of ARDS, although care must be taken to avoid the effects of high MAP on ICP. In patients requiring high levels of PEEP, the head of the bed should be raised to minimize the effect of the increased intrathoracic pressure and ICP should be carefully monitored.

The choice of volume-controlled or pressure-controlled ventilation is based on clinician's bias. A tidal volume in the range of 6 to 8 mL/kg ideal body weight can be used provided that plateau pressure is kept below 30 cm H₂O. This is usually not a problem, because these patients typically have a nearly normal lung and chest wall compliance. The ventilatory goal is to maintain the Pco₂ 35 to 45 mm Hg and pH 7.35 to 7.45. If the patient has concomitant acute or chronic respiratory disease, a lower tidal volume is selected. A respiratory rate appropriate to achieve normal acid-base balance should be chosen. This can often be achieved at a rate of 15 to 25 breaths/min. An inspiratory time of 1 second is usually adequate.

Monitoring

Monitoring of mechanically ventilated head-injured patients is similar to that of any mechanically ventilated patient (Table 20-4). If minute ventilation is increased to produce iatrogenic hyperventilation for a short period of time, the presence of auto-PEEP must be evaluated. Capnography may be useful to monitor the level of ventilation in these patients, who often have normal lung function and do not tolerate well an increase in Paco₂.

Close observation of ICP should be used when ventilator settings are manipulated. If an ICP monitor is not present, clinical signs of an increased ICP (eg, pupillary response, posturing, changes in level of consciousness) should be evaluated when ventilator changes occur. Although airway clearance is important in these patients, care must be taken to avoid deleterious increases in ICP during suctioning. Nutritional support is necessary to facilitate healing and weaning from mechanical ventilation. Pulmonary embolism can occur in patients with prolonged immobility, and pulmonary infection is also common in these patients.

Jugular venous bulb oxygen saturation (Sjvo₂) and brain Po₂ (P_bo₂) from a probe placed into the brain may be used as an index of the adequacy of cerebral oxygenation. The use of these monitors is controversial. If used, Sjvo₂ less than 50% or P_bo₂ less than 15 mm Hg are treatment thresholds.

Liberation

Liberation should not be considered until respiratory depressant therapy is no longer required. Ventilator discontinuation can often be initiated before the patient's neurologic function is maximally restored if the ventilatory drive is intact. For some patients, maintenance of a stable airway is required for a longer time than ventilatory support (ie, tracheostomy). However, extubation should not be delayed solely on the basis of depressed neurologic status. Due to central neurologic dysfunction, ventilator liberation, extubation, and decannulation of some of these patients can be prolonged and difficult. Weaning approaches should incorporate spontaneous breathing trials and appropriate rest following a failed trial of spontaneous breathing.

Apnea Test

An apnea test is commonly conducted as part of the diagnosis of brain death. Before conducting the apnea test, the following prerequisites should be met: core temperature more than or equal to 36.5°C, systolic blood pressure more than or equal to 90 mm Hg, euvolemia, normoxemia (or Pao₂ > 200 mm Hg breathing 100% oxygen), and eucapnia (or Paco₂ > 40 mm Hg in the patient with chronic hypercapnia). The following procedure is used:

• Disconnect the ventilator.

• Administer 6 L/min O₂, either by T-piece or by a catheter passed into the trachea.

• Observe the patient closely for signs of respiratory movements. If respiratory movements occur, the apnea test is negative (ie, does not support the clinical diagnosis of brain death), and mechanical ventilation is resumed.

• If respiratory movements do not occur, measure arterial blood gases after 8 minutes and reconnect the ventilator.

• If respiratory movements are absent and Paco₂ is more than 60 mm Hg (or 20 mm Hg greater than baseline), the apnea test result is positive and consistent with the clinical diagnosis of brain death.

• If hypotension or desaturation occurs during the apnea test, the ventilator is reconnected and the test is resumed at a later time.

• If no respiratory movements are observed, Paco₂ is less than 60 mm Hg, and no adverse effects occur, the test may be repeated with 10 minutes of apnea.