Evolving Injury and Repeat Head CT

Progressive intracranial hemorrhage consistent with an evolving contusion is seen in 14% to 38% of patients (57,78,79). Although worsening CT findings does not necessarily require treatment, up to 54% of patients may require neurosurgical intervention including ICP monitoring or craniotomy subsequent to the findings on a repeat scan (53,80–85). A significant risk factor is early initial imaging within 2 hours of injury (54,78,85,86). Often community standards are to repeat a CT scan within 12 to 24 hours of the initial imaging.

In stable patients without clinical neurologic deterioration, the utility of repeat imaging is debated and many believe it is unnecessary since it is unlikely that neurosurgical intervention will be required (54,82,85,86). Other independent risk factors for progression include associated tSAH, SDH, older age, and prolonged partial thromboplastin (53,57,78,85,86). A large initial contusional or IPH size and effacement of cisterns are strongly predictive of failure of nonoperative management (57).

Hypoxemia and Respiratory Management

Hypoxemia, defined as a PaO₂ less than 60 mmHg or O₂ saturation less than 90%, can independently increase mortality from 27% to 50% and increase poor outcome from 28% to 71% (87,88). Early intubation can prevent aspiration and minimize hypoxic and hypercapnic events (89), and is recommended by the Eastern Association for the Surgery of Trauma (EAST) (90), Advanced Trauma Life Support (ATLS) guidelines from the American College of Surgeons (91), and the Brain Trauma Foundation Traumatic Brain Injury prehospital guidelines (92). A GCS of 8 or less is the usual threshold for endotracheal intubation.

There are caveats to intubation of which the practitioner should be aware. In the prehospital setting, rapid sequence intubation has been associated with increased mortality (93–95). This may be a result of decreased cerebral perfusion due to hyperventilation-induced hypocapnia. Positive pressure ventilation may cause hypotension in a hypovolemic patient if central venous return is impeded by high intrathoracic pressures (96). Intubation is a high-risk procedure which may cause secondary neurologic injury. Sedative/hypnotic medications and bag/mask ventilation with positive pressure ventilation contribute to hypotension and hypercapnia and hypocapnia, respectively, during induction. In addition, direct laryngoscopy causes a marked, transient increase in ICP. Intravenous lidocaine may blunt this ICP response (97,98).

Hyperventilation with resultant hypocapnia causes cerebral vasoconstriction and a reduction in CBF (99–101). Prolonged hypocapnia appears to slow neurologic recovery (102). Prophylactic hyperventilation of PaCO₂ less than 35 mmHg should be avoided, although PaCO₂ as low as 30 mmHg may be necessary for brief periods for immediate treatment of intracranial hypertension. Options to identify cerebral ischemia in the setting of hyperventilation include the use of jugular venous oxygen saturation, arterial jugular venous oxygen content differences, brain tissue oxygen monitoring or CBF monitoring (103).

To achieve adequate ventilation, positive end-expiratory pressure (PEEP) may be necessary. PEEP affects CPP and ICP when the lung is compliant and the chest wall is not. The high lung compliance allows for an increased intrathoracic volume which, in the setting of a low compliant chest wall, increases intrathoracic pressures. The high intrathoracic pressure decreases cerebral venous outflow, which will increase ICP (104,105). Additionally, if intrathoracic pressure is elevated, cardiac venous return is diminished and results in lowered MAP and CPP.

Pulmonary infections were seen in 41% of patients registered in the TCDB and were an independent predictor of unfavorable outcome (106). Bedside management includes adequate pulmonary toilet and strategies such as elevation of the HOB to decrease the risk for ventilator-associated pneumonia (VAP). In patients with intracranial hypertension, during endotracheal suctioning, adequate sedation is necessary to prevent an increase in ICP (107).

Neurogenic Pulmonary Edema

In addition to hypoventilation and aspiration from poor airway protection, a less frequently recognized cause of hypoxemia following TBI is neurogenic pulmonary edema (NPE) (108), resulting from central sympathetic stimulation. Pretreatment with adrenergic-blocking agents prevents experimental NPE (109). Experimental lesions in the hypothalamus (110), bilateral nucleus tractus solitarius (111), and the ventrolateral medulla (112) can produce NPE. TBI causes a sympathetic discharge, which increases systemic and pulmonary vascular pressures; the resultant increase in pulmonary capillary pressure increases the hydrostatic pressure and causes pulmonary capillary injury. This, in turn, causes leakage of fluid and protein and pulmonary hemorrhages (113–115).

Clinical signs include dyspnea, tachypnea, tachycardia, and chest pain if the patient is awake; rales are present on chest auscultation. Laboratory results show hypoxemia and a mild leukocytosis, with chest radiograph showing a bilateral alveolar filling process (116). Pulmonary capillary wedge pressures and pulmonary artery pressures can be elevated or normal. There are two distinct forms of NPE: the classic form appears early, within minutes to a few hours after acute brain injury; a delayed form of NPE slowly progresses over 12 to 72 hours following injury. Treatment is supportive and often requires supplemental oxygen and positive pressure ventilation. Dobutamine may be effective by decreasing cardiac afterload and increasing cardiac contractility (117).

Hypotension

Hypotension, defined as SBP less than 90 mmHg, independently worsens mortality (118,119). The TCDB reports hypotension was present in 29% of patients and doubled mortality from 27% to 55% (106,118). In patients whose SBP were less than 90 mm Hg, mortality was 65% independent of age, admission GCS motor score, hypoxia, or associated severe extracranial trauma. Adequate fluid resuscitation with euvolemia is essential. Independent of ICP, MAP, or CPP, a negative fluid balance of approximately 600 mL was associated with poorer outcome (120). Guidelines recommend adequate fluid resuscitation and have been updated to recommend SBP 100 mm Hg or greater in patients 50 to 69 years old and 110 mm Hg or greater in patients 15 to 49 or greater than 70 years old. Blood pressure support to maintain SBP greater than 90 mmHg (67). Once an ICP monitor is placed, the optimal blood pressure is determined by that required to keep the CPP 60 to 70 mmHg (67).

Contraction Band Necrosis

Following head trauma, SAH, seizures, or stroke, patients may have cardiogenic shock with global hypokinesis associated with transient cardiac dysrhythmias and repolarization changes (121–124). Dysrhythmias may include supraventricular tachycardias, sinus bradycardia, atrioventricular (AV) block, AV dissociation, nodal rhythms, and paroxysmal ventricular tachycardia. These changes are cerebrally mediated and are recognized as myofibrillar degeneration (also known as contraction band necrosis [CBN] or coagulative myocytolysis). The histologic appearance of CBN contrasts to the coagulation necrosis seen with ischemic injury where there are cytoplasmic degenerative changes with cloudy swelling, hyaline droplets, and fatty change. With CBN, the myocardium instead shows loss of definition of the linear arrangement of myofibrils and the appearance of prominent dense eosinophilic transverse bands (contraction bands), and intervening granularity throughout the cytoplasm (125). This injury pattern was first described with pheochromocytoma and has been associated with the administration of catecholamines, including cocaine abuse (126,127). It is postulated that centrally mediated sympathetic or exogenous catecholamine stimulation of the myocardium results in cellular calcium overload and results in the formation of the contraction bands (128). CBN is predominantly located in the subendocardium with the cardiac conducting system, which results in the associated arrhythmias (129).

In CBN, cardiac enzymes are often elevated and may be difficult to differentiate from an acute coronary syndrome. However, the treatment for CBN is vastly different and typically includes observation for dysrhythmias and blood pressure support in contrast to reperfusion therapy with an acute ischemic myocardial infarction. Clinical differentiation typically relies on the recognition of patients with higher risk for coronary artery disease such as older age, hypertension, diabetes, and hyperlipidemia rather than a young patient with massive head injury.

Posttraumatic Vasospasm

Following TBI, posttraumatic vasospasm can occur in as many as 24% to 36% of adults and children (130–132) and causes focal ischemia with lateralizing neurologic deficits such as hemiparesis and aphasia between 2 to 37 days following injury (132–134). In patients with severe TBI, small studies have reported incidences as high as 82% (135). Explosive blast TBI is especially associated with early cerebral edema and cerebral vasospasm (136). Disruption of cerebrovascular tone may be a result of inflammatory and other vascular changes (137). Some mechanisms include an increased expression of an inducible isoform of nitric oxide synthase (138) and a hypercontraction-induced phenotypic switch that potentiates vascular remodeling (139). Transcranial Doppler, while reasonably specific, is not a sensitive test for vasospasm. If vasospasm is suspected, cerebral angiography can confirm the diagnosis. The effectiveness of treatment of posttraumatic vasospasm with modalities used following aSAH (e.g., hypervolemic, hypertensive therapy, or nimodipine) has not been assessed.

Hyperthermia

Hyperthermia accelerates neuronal injury by increasing basal energy requirements (neuronal discharges), excitatory neurotransmitters, free radial production, calcium-dependent protein phosphorylation, ICAM-1 and inflammatory responses, DNA fragmentation, and apoptosis, causing blood–brain barrier changes as seen by extravasation of protein tracers (140,141). Despite this, multiple TBI studies of prophylactic moderate hypothermia (32° to 33°C) and their meta-analyses have not shown improved outcome (142–144). This may be due to significant intercenter variability in the management of MAP, CPP, fluids, and vasopressors (145,146).

In the individual patient, therapeutic hypothermia lowers ICP by reducing the cerebral metabolic rate 7% for each degree Celsius decrease. This treatment can be lifesaving and result in reasonable neurologic recovery (144,147). Pentobarbital coma and/or neuromuscular blockade (NMB) may be necessary to achieve cooling without shivering. Various techniques for intravascular and topical cooling are available. Although complications of hypothermia can include increased risk of cardiac dysrhythmias, hypotension, bradycardia, thrombocytopenia, and pneumonia, in studies evaluating hypothermia in cardiac arrest patients, there was no statistical increase in these adverse events (148,149).

Hyperglycemia

Hyperglycemia can cause brain tissue acidosis (150), and early hyperglycemia has been associated with worsened neurologic outcome following TBI (151,152). Persistent hyperglycemia following severe TBI in one study was an independent risk factor of mortality with a 4.9 times increase in risk of death (153). It is not fully understood whether the hyperglycemia is causative or is a marker for severity of injury and subsequent poor outcome; however, control of hyperglycemia following TBI is theoretically reasonable.

In the ICU setting, where glycemic control often uses insulin infusion or injection, patients with acutely altered mental status should be urgently evaluated for hypoglycemia.

Coagulopathy

Brain is rich in tissue thromboplastin and following head injury, increased tissue thromboplastin activity in the frontal, parietal, and temporal lobes activates the coagulation cascade and causes a disseminated intravascular coagulopathy (DIC) (154). The TCDB reported that 19% of patients were coagulopathic (106). In children, the incidence of coagulopathy increased with worsening severity of head injury reflected by the head Abbreviated Injury Scale score and was as high as 40% (155). Although initial evaluation may show thrombocytopenia in 14% and coagulopathy in 21% of TBI patients, in ensuing days, DIC can be seen in 41% to 60% of patients with blunt brain injury (156,157). It is more common in patients with penetrating head trauma (158).

Abnormalities in PT, PTT, or platelet count have been associated with 55% of patients who have progressive intracranial hemorrhage after TBI (159,160). Associated coagulopathy and thrombocytopenia increases mortality in TBI (156,157,160,161). Some centers are using thromboelastometry and portable coagulometers to detect coagulopathy in the ED. Typically, fresh frozen plasma is transfused for an elevated aPTT or a PT international ratio (INR) of 1.5 or more. Current European guidelines for TBI patients recommend transfusion of platelets for values below 100,000 cells/μL or for patients receiving antiplatelet agents (162). Other alternatives such as activated factor VII or prothrombin complex concentrate (PCC) may be effective emergently (158); PCC is recommended in the European guidelines for emergency reversal of vitamin K-dependent and oral antifactor Xa agents such as rivaroxaban, apixaban, or edoxaban (162).

Intracranial Pressure Monitoring and Management

Normal ICP is less than 10 mmHg; the TCDB reports that 72% of patients with severe TBI had ICPs above 20 mmHg (163). Since multiple studies show worsened outcome with ICP above 20 to 25 mmHg, published guidelines use this as the threshold to treat (65).

Maneuvers for management of ICP begin with those with fewer potential side effects and progress to more invasive treatments with higher complication risk (Fig. 117.1). Elevation of the head of bed to more than 30 to 45 degrees not only decreases the risk of VAP but can facilitate cerebral venous drainage and lower ICP. In orthostatic, hypovolemic patients, however, head of bed elevation can lower MAP; adequate fluid resuscitation is necessary. Adequate pain therapy with opioids and adequate sedation with sedative-hypnotics decrease ICP. Constant infusion may be hemodynamically better tolerated than bolus administration. Prophylactic or sustained hyperventilation of PaCO₂ less than 35 mmHg may be harmful and should be avoided (101–104). In the situation of impending herniation or refractory intracranial hypertension, decreasing PaCO₂ to 30 mmHg for transient “rescue” therapy will give the practitioner time to initiate other maneuvers to lower ICP.

Osmotic therapy is a mainstay in ICP management. Mannitol and hypertonic saline are both effective (164–166). However, repetitive dosing by the above agents may eventually shift fluid into the injured brain across the damaged blood–brain barrier and thereby increase the volume of injured brain and elevate ICP (167–169). To avoid this, osmotic therapy, similar to hyperventilation, might be best used as rescue therapy until more definitive therapy is implemented. Doses of 0.25 to 1 g/kg of mannitol are effective. The lower dose drops ICP and may decrease the risk of vasogenic edema seen with multiple dosing (167). High serum concentrations of mannitol may cause renal failure. Maximal mannitol dosing traditionally has been when serum osmolarity reaches 320 mOsm/kg, but with the increased use of hypertonic saline, an osmolar gap is now being used. The difference of measured to calculated serum osmolarity is a surrogate marker for mannitol concentration as an “unmeasured” osmole. A rising osmolar gap of greater than 10 from baseline may be a better indicator of maximized mannitol administration (170,171). Hypertonic saline may be more effective and have a longer duration of action on lowering ICP than mannitol (169,172,173). Usual doses include 250 mL of 3% or 30 mL of 23% saline which have equimolar amounts of sodium chloride (174). The concentration used usually depends on volume status, with the lower volume 23% saline used when the patient is hypervolemic. Osmotic therapy is best administered through a central venous access as it may sclerose veins. When deciding which osmotic agent to use, elevated ICP with low fluid status would be best treated with hypertonic saline. Studies evaluating the prophylactic use of hypertonic saline compared to conventional fluids for prehospital or ED resuscitation have not shown outcome improvement (175,176).

NMB lowers ICP by decreasing muscle tone, especially during shivering. Shivering increases the metabolic rate and generates carbon dioxide. After administering NMB, an ABG analysis should be obtained to ensure that the PaCO₂ has not dropped below 35 mmHg; minute ventilation should be adjusted accordingly. Rapid increases in PaCO₂ may result in rebound vasodilation and ICP elevation. Ventilator manipulation should be performed in small increments when adjusting to increase the PaCO₂. NMB also assists in cooling the patient. As noted above, hypothermia is useful in refractory intracranial hypertension. Temperatures of 32° to 33°C can be well tolerated. On a cautionary note, the combination of NMB and cooling appears to put the patient at high risk for pneumonia, as they are unable to effectively clear secretions; empiric pulmonary toilet with frequent suctioning is often necessary.

Barbiturate-induced coma, using either thiopental or pentobarbital infusion, lowers the cerebral metabolic rate; this results in lowered cerebral blood volume and ICP. Thiopental in long-term infusion, because of its lipophilicity, may take over a week to clear after the infusion is stopped. For pentobarbital, 20 mg/kg is given as a slow loading dose followed by a maintenance infusion of 1 mg/kg/hr; the loading dose may significantly lower MAP. Often fluids and vasopressor administration may be necessary. Electroencephalography (EEG) is critical to titrating the dose during barbiturate coma. Although the infusion can be titrated to ICP effect, once the EEG is isoelectric, there is little to be gained in the way of ICP control by increasing the infusion; at this point, worsening side effects result from increased drug. These include hypotension from peripheral vasodilation, decreased cardiac inotropy, and ileus; cough reflex is diminished and decreased bronchociliary activity and slowed leukocyte chemotaxis increase the risk for pneumonia. A benefit of barbiturate coma is a quiescent hypothalamus that no longer modulates body temperature. Hypothermia can often be achieved without the need for NMB since shivering is diminished.

Loop diuretics have been used to help manage ICP by decreasing CSF production in the choroids plexus (177). Loop diuretics will decrease volume status, thus unless the patient is hypervolemic, CSF diversion is a more effective method of lowering ICP.

Other therapy for refractory intracranial hypertension requires neurosurgical intervention. Placement of an EVD allows for CSF drainage. Hemicraniectomy may be life-saving and a viable option depending on the patient; case series of 19, 23, and 51 children at three different centers had mortalities of 30% to 31.4%. Favorable outcome with return to school and functional independence was reported in 68% to 81%; 18% to 21% were severely disabled and dependent on caregivers (178–180). In a randomized trial of adults with diffuse brain injury, early bilateral frontotemporoparietal decompressive craniectomy decreased ICP and the ICU length of stay, but was associated with worsened neurologic outcome (181).

Brain Tissue Oxygenation

To monitor and help prevent the secondary damage seen with hypoxic brain injury, new modalities to evaluate cerebral oxygenation have been developed. Jugular bulb oximetry (SjvO₂) is a global measure of the balance between oxygen delivery to the brain and oxygen consumption. Local brain tissue partial pressure oxygen (PbtO₂) can be measured either by a quenching process by fluorescence or by a polarographic Clark-type microcatheter. An increase in cerebral oxygen delivery is reflected by increases in SjvO₂ and PbtO₂. Oxygen delivery to the brain is manipulated by increases in blood pressure, cardiac output, and red blood cell transfusion (182). Normobaric hyperoxia has not shown to improve cerebral oxygen metabolism on PET imaging, and the use of 100% oxygen is not supported by the available literature (183). Optimal SjvO₂ is generally accepted as 50% oxygen saturation (103,184). The optimal PbtO₂ is not fully established but guidelines currently recommend higher than 15 mmHg (103). Various studies show worsened outcome in patients with mean PbtO₂ less than 15 mmHg; other thresholds include 25 mmHg (185–188). Mortality was significantly decreased and functional outcomes improved in one study comparing 25 patients treated by traditional ICP/CPP-guided therapy to 28 patients with therapy targeted to a PbtO₂ greater than 25 mmHg (189); however, other studies have not shown improvement in outcome (190–192). Cerebral microdialysis evaluating the biochemical byproducts of ischemia such as increased lactate and glutamate and lactate–pyruvate ratio is another potential technology to assist bedside care, but has not yet reached practical clinical use (193).

At the time of this writing, an evaluation of 31 adult neurocritical care units in the United Kingdom managing TBI patients showed that 100% of units followed ICP; 97% monitored CPP with 25 of 31 units using a CPP target of 60 to 70 mmHg; PbtO₂ was utilized in 26% of units, cerebral microdialysis in 13%, and SjvO₂ was used in only one unit (194). No unit was using near-infrared spectroscopy, a modality developed to measure cerebral oximetry as a measure of perfusion.

Antibiotic Prophylaxis

Fractures of the skull base and severe facial trauma can result in a CSF leak. Various studies report incidences of 2.6% to 4.6% of all patients with basilar or facial fractures (195,196). In one study, otorrhea was three times more common than rhinorrhea (195). Approximately 50% of CSF leaks stop within 5 days (197); the risk of bacterial meningitis is approximately 12% to 21%. Studies conflict as to whether prophylactic antibiotics decrease the risk of infection and meta-analyses suggest no benefit, hence, there are no guidelines or recommendations regarding antibiotics in this setting (197–199); constant surveillance for meningitis is essential.

In the setting of CSF leak, if the spine is stable and blood pressure is adequate, the HOB should be elevated to facilitate leak closure. Stool softeners help avoid vigorous Valsalva maneuvers that may worsen the leak. Neurosurgical intervention with CSF diversion (i.e., lumbar drain or EVD) or surgical closure may be necessary if the leak persists. Following penetrating head trauma, a CSF leak is the primary predictor of intracranial infection. Infection is seen in 38% to 63% of CSF leaks after military-related penetrating cerebral injury (200–202). Current recommendations advise treatment for 5 to 14 days with empiric broad-spectrum antibiotics immediately following penetrating brain injury (202–205).

For clean neurosurgical procedures, such as twist drill craniostomy for EVD or ICP monitor placement, burr holes, or craniotomy, guidelines have been established by the Surgical Infection Prevention and Surgical Care Improvement Projects that recommend an intravenous first-generation cephalosporin within 1 hour prior to surgical incision (206).

Posttraumatic Seizures

Early posttraumatic seizures occur within 7 days of injury; 3% to 6% of patients with closed head injury suffer early posttraumatic seizures, compared to 8% to 10% with penetrating brain injury (207–209). Late posttraumatic seizure by definition manifests at least 7 days postinjury and is seen in 30% of patients with penetrating brain injury; these late posttraumatic seizures may occur up to 5 years after injury. There is adequate evidence to recommend antiseizure medications, for example, phenytoin and carbamazepine, for the first week after closed and penetrating brain injury to prevent early posttraumatic seizures (210–213). Of note, valproate showed no benefit for seizures following brain injury and had a trend to higher mortality (214). Levetiracetam, compared to phenytoin, shows equal efficacy in post-TBI seizure prevention (215,216) and has fewer complications, such as hypotension, and does not require monitoring of therapeutic levels, although dosing should be adjusted in renal failure. There is no evidence that continuing prophylactic antiseizure medications beyond a week prevents late posttraumatic seizures, and it is not recommended for closed or penetrating head injury beyond 7 days of injury (210,213).

Thromboprophylaxis

In the general postoperative neurosurgical population, the risk for deep venous thrombosis (DVT) is 3% to 14% (217–220); following major head injury, the risk for DVT is as high as 54% (221). The BTF recommends the use of graduated compression stockings or intermittent pneumatic compression (IPC) stockings in combination with low–molecular-weight (LMWH) or low-dose unfractionated heparin (LDUH) with the warning that this may result in intracranial hemorrhage expansion; there were no specific recommendations for a preferred agent, dose, or timing of treatment (222). The American College of Chest Physicians (ACCP) 9th edition has similar recommendations for major trauma patients with TBI, ASCI, or spinal surgery for trauma (Grade 2C). The Neurocritical Care Society recommends initiating LMWH or UFH for VTE prophylaxis within 24-48 hours of presentation or 24 hours after craniotomy (223). If LMWH and LDUH are considered contraindicated, mechanical prophylaxis, preferably with IPC devices, should be utilized until the risk of bleeding diminishes and contraindication to chemoprophylaxis resolves (Grade 2C). Inferior vena cava (IVC) filters are not considered acceptable primary prophylaxis in trauma patients (Grade 2C) (220).

Nutrition

Following severe TBI, patients enter a hypermetabolic, catabolic state with rapid weight loss associated with a negative nitrogen balance and protein wasting. In experimental models of TBI, 3 hours after injury morphologic changes are seen in the gut mucosa that include shedding of epithelial cells, fracture of villi, focal ulcers, fusion of adjacent villi, mucosal atrophy, and edema in the villous interstitium and lamina propria. On electron microscopy, there is a loss of tight junctions between enterocytes, damage of mitochondria and endoplasm, and apoptosis of epithelial cells (224). These changes in gut permeability increase bacteria translocation and endotoxin, which increases the risk of the systemic inflammatory response; the amino acids arginine and glutamine modulate gut permeability.

Early parenteral or enteral nutrition within 24 to 72 hours can speed neurologic recovery and decrease disability and mortality (225–228). Early enteral feeding may have benefit over parenteral feeding by protecting against intestinal apoptosis and atrophy (229) and decreasing infection clinically (230). Early enteral nutrition with glutamine and probiotics may decrease the infection rate and length of ICU stay (231,232). There is some theoretical concern that glutamine should not be used in brain injury patients due to the potential increase in cerebral glutamate with neuroexcitatory properties and cell damage, although there are no data to date to support this concern.

Fasted TBI patients lose nitrogen at a rate that reduces weight by 15% per week. Replacement of resting energy expenditure (REE) by 100% to 140%, with 15% to 20% nitrogen calories may reduce nitrogen loss; therefore, BTF guidelines recommend that full (100% REE) nutritional replacement be achieved within 7 days after injury (233). An Institute of Medicine (IOM) report was more aggressive and recommended the provision of early (i.e., within 24 hours of injury) nutrition of more than 50% of total energy expenditure and 1 to 1.5 g/kg protein for the first 2 weeks after injury (232,234).

Stress Gastritis

Stress gastritis was seen in 91% of 44 comatose mechanically ventilated patients within 24 hours of head injury; lesions were most commonly seen in the fundus and body of the stomach (235). Mucosal ulceration is typically prevented by maintaining intraluminal pH above 5 or by H₂ receptor blockade (236). In TBI patients, stress ulcer bleeding prophylaxis with proton pump inhibitors or H₂ antagonists is recommended. Neither is recommended over the other, but the practitioner should be aware that one retrospective study in neurosurgery patients showed a statistically significant increased incidence of thrombocytopenia of 50 patients on famotidine compared to 98 of those not treated (34% vs. 11.2%) (237).

Prognosis

Survivors of TBI variably suffer from long-term cognitive, motor, sensory, and emotional deficits. These manifest as weakness, incoordination, emotional lability, impulsivity, and difficulty with vision, concentration, memory, judgment, and mood. Nearly 5.3 million in the United States live with disabilities as a result of TBI (14). When the postresuscitation GCS is not complicated by medications or intubation, approximately 20% of patients with GCS 3 will survive and 8% to 10% will have moderate to good recovery such that they are able to live independently (238). Despite this, 34% to 47% of “minor” head injury patients cannot return to work or their previous lifestyle (239,240). Independent predictors of outcome include older age at time of injury, the postresuscitation Glasgow coma score, injury severity score (ISS), pupillary response on admission, and CT scan findings of diffuse edema, tSAH, SDH, partial obliteration of the basal cisterns, or midline shift (241–246). The TCDB reports a postresuscitation mortality rate of severe TBI patients of 76% and 18%, for patients with postresuscitation GCS of 3 and 6 to 8, respectively. Overall mortality was 36% in 746 patients (42). In another study of 1,311 head-injured patients, the highest mortality was associated with spinal cord injury, obstructed airway, difficulty breathing, and shock, although none of these was independently predictive of survival when adjusted for GCS (247). There are on-going trials to find medications to improve outcome following TBI; studies to date, with agents such as free radical scavengers (248) and progesterone (249) have been unsuccessful.

Pathophysiology

Primary spinal cord injury results from cord compression from discs, bone, ligament, or hematoma, or from distractional forces such as flexion, extension, dislocation, or rotation, which cause shearing of the neuronal axons or vasculature with intramedullary hemorrhage. Similar to head injury, the spinal cord undergoes both primary and secondary injury. Secondary injury can result from additional mechanical injury if the spine is manipulated when it is unstable or as a result from systemic and local vascular insults, which may be a result of hypotension, electrolytes changes, edema, and excitotoxicity (250).

Diagnosis

Examination

The physical examination includes a general examination to survey for other injuries; the quality of the patient’s breathing should be assessed to ensure adequate ventilatory effort. The neurologic examination should include a mental status evaluation as concomitant head injury may occur. A complete cranial nerve examination assesses for evidence of cranial neuropathies or nystagmus suggestive of brainstem or cerebellar ischemia that may result from vertebral artery injury. In 1982, the American Spinal Injury Association (ASIA) developed the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) (Fig. 117.2) (251) to improve precision in determining the level and extent of neurologic injury for the National SCI Statistical Center Database (252). These have been updated since that time, have good interrater reliability (253,254), and are recommended as the preferred neurologic examination tool (255). The ASIA Impairment Scale standardizes language used to describe severity SCI (Table 117.2) (251). The single neurologic level of injury is defined as the most caudal segment of the cord with intact sensation and antigravity muscle function strength.

Radiologic Evaluation

No cervical radiologic evaluation is recommended in an awake, alert, nonintoxicated trauma patient who has no neck pain or tenderness and is neurologically normal unless there are significant associated injuries that would interfere with the history and physical examination. In this setting, cervical immobilization can be discontinued without imaging. However, in patients with neck pain or tenderness, a high-quality cervical spine CT should be performed. The CT scan is better than MRI for evaluating bones; however, for ligamentous injury, dynamic flexion/extension radiographs in an awake patient are preferred. Alternatively, an MRI within 48 hours of injury can also detect ligamentous injury (256). Newer-generation CT scanners are sensitive, and routine three-view (anteroposterior, lateral, and odontoid) cervical spine radiographs are not recommended unless high-quality CT imaging is not available. In obtunded patients, again, high-quality cervical CT is recommended to rule out clinically significant injury (257). However caution is taken before removing and immobilizing collar since an obtunded patient cannot report pain associated with ligamentous injury and instability. Options to remove the collar include: (1) waiting until the patient is awake and asymptomatic; (2) a normal MRI obtained within 48 hours of injury (although this is considered level III evidence with limited or conflicting data); (3) at the discretion of the treating physician (256) or 4) more recent guidelines recommend the removal of the collar following a normal high quality CT scan defined as 3 mm slices or less (258).

TABLE 117.2 American Spinal Injury Association Impairment Scale

MRI is recommended for patients with cervical fracture-dislocation injuries if they cannot be examined during closed reduction. An MRI is performed to evaluate for disrupted or herniated intervertebral discs, which are found in up to 33% to 50% of patients with facet subluxation injuries (259). MRI is also recommended for patients with occipital condyle fractures to assess the integrity of the craniocervical ligaments (260).

Treatment

Associated Vascular Injury

Blunt cervical spinal trauma can result in vascular injury and cause cerebral ischemia (291). Incidence varies from 0.03% to 4.8% and likely depends on the screening method (292–294). Mortality ranges from 23% to 28%, while 48% to 58% of survivors have significant neurologic deficits (295). The most common mechanism of blunt cerebrovascular injury (BCVI) is MVC, followed by falls, and pedestrian and motorcycle crashes. Many institutions use modifications of the Denver Screening Criteria for BCVI based on risk factors to determine whom to screen (296). Risk factors include cervical fractures with subluxation or with a fracture through the transverse foramen, displaced or complex midface or mandibular fracture (LeForte II or III), a basilar skull fracture involving the carotid canal or sphenoid sinus, near-hanging resulting in cerebral hypoxia, and cervical vertebral body fraction or distraction injury, a seatbelt sign or tissue injury of the anterior neck, and massive epistaxis or a cervical hematoma (292). Suspicion should be high if the patient develops a lateralizing neurologic deficit with an initially normal CT scan, or evidence of a recent ischemic stroke on cerebral imaging. CT angiography (CTA) is sensitive for BCVI and has taken the place of conventional cerebral angiography as the recommended screening tool unless CTA is not available or endovascular treatment is anticipated (297–300). In a patient with a vertebral subluxation or complete spinal cord injury, MRI is the recommended diagnostic modality (300).

A grading system of injury has been described by Biffl (301) (Table 117.3). Fifty-seven percentage of grade I injuries heal spontaneously in 10 days independent of therapy (302); therefore, they can be treated with aspirin. Retrospective studies of grade II through IV injuries show no difference between antiplatelet agent or heparin therapy although heparinization increases hemorrhage risk (303,304). Current recommendations suggest individualized therapy dependent on the vascular injury, associated injuries, and the risk of hemorrhage. Options include no treatment, antiplatelet therapy, or anticoagulation. No recommendations regarding endovascular therapy can yet be made (300).

Thromboprophylaxis

DVT detection with ¹³¹I-fibrinogen scans of patients with ASCI and paralysis ranges as high as 100% (305). Recommended diagnostic tests include: duplex Doppler ultrasound, impedance plethysmography, venous occlusion plethysmography, venography, and the clinical examination (306).

TABLE 117.3 Grading Scale for Blunt Cervical Vascular Injury

Related posts:

Policy, Politics, and the Intensive Care Unit Understanding Reactions of Patients and Families and Sharing Difficult News Pleural Disease in the Intensive Care Unit The Geriatric Patient Post-ICU Syndrome and the Post-ICU Clinic Neuromuscular Disorders