Trauma and Acute Care

Emergent-Urgent-Nonurgent

Table 17-1 is an outline of trauma case priorities. Emergent cases must reach the OR as soon as possible. Although surgical airway access and resuscitative thoracotomy usually occur in the ED, immediate follow-up in the OR will be necessary if the patient survives. Also considered emergent are any exploratory surgeries (laparotomy or thoracotomy) in a hemodynamically unstable patient and craniotomy in a patient with a depressed or deteriorating mental status, when evacuation of blood or decompression of severe cerebral edema will result in a survival benefit. Limb-threatening orthopedic and vascular injuries should undergo surgical exploration as soon as the necessary diagnostic studies have been performed and interpreted.

Urgent cases are not immediately life threatening but require surgery as soon as possible to reduce the incidence of subsequent complications. Examples include exploratory laparotomy in stable patients with free abdominal fluid; irrigation, debridement, and initial stabilization of open fractures; and repair of contained rupture of the thoracic aorta. Angiographic procedures have increasingly replaced open surgeries for splenic, hepatic, pelvic, and aortic injuries in hemodynamically stable patients. Early fixation of closed fractures, especially spine and long-bone fractures, has been shown to benefit trauma patients by reducing the incidence of subsequent pulmonary complications. Definitive repair within 24 hours is recommended in otherwise stable and non–brain-injured patients.

Nonurgent cases are those that can be safely delayed until a scheduled OR time is available. Although immediate fixation of face, wrist, and ankle fractures may shorten the patient's length of stay, early surgeries may be technically more difficult because of swelling and distortion of the surrounding tissue. Therefore, such procedures are typically postponed days to weeks after injury, when tissue edema has resolved and the patient's condition is improved. Early pain control is critical to mitigate the inflammatory response and development of long-term pain syndromes.

Damage Control Approach

In addition to facilitating timely surgery in patients who require it, the anesthesiologist, surgeon, and other specialists work together to determine the extent of surgery allowed by the patient's physiology. The concept of “damage control” has revolutionized surgical thinking in the last two decades, limiting initial therapeutic procedures only to those required for the achievement of hemostasis and homeostasis, while delaying reconstructive procedures until adequate resuscitation has been achieved, and in appropriate cases, edema has subsided. In a typical example, the surgeon treating an unstable patient with blunt trauma might perform an exploratory laparotomy, rapid splenectomy, staple resection of injured bowel (without attempt at reanastomosis), ligation of bleeding large vessels, and packing of the abdomen. The abdomen would be left open under a sterile, watertight wound vacuum and the patient taken to the ICU. Angiographic embolization might be necessary to facilitate hemostasis in the liver and retroperitoneum (e.g., because of pelvic fractures). The goal with “damage control” surgery is to avoid the “lethal triad” of hemorrhage, acidosis, and coagulopathy that can rapidly develop in a patient with massive bleeding and resuscitation. After resolution of shock, warming, and normalization of laboratory values, the patient would return serially to the OR in 24 to 48 hours for further exploration and debridement of nonviable tissue, reconstruction of the bowel, placement of enteral feeding access, and abdominal closure.

The concept of damage control may also be applied to orthopedic injuries: initial external fixation of the pelvis and long bones is adequate for temporary stabilization of fractures, without assuming the additional physiologic risks of intramedullary nailing or open fixation. The damage control approach should be considered in any patient with persistent hypoperfusion, elevated lactate, or transfusion requirement in excess of one blood volume.

Pathophysiology

Baseline indications for intubation of the trauma patient are similar to those of any critically ill patient and can be organized under basic categories of (1) inability to oxygenate, (2) inability to ventilate, and (3) inability to maintain a patent airway. Indications may also include need for pain control (multiple fractures), diagnostic workup, or plan to proceed to the OR. Box 17-1 lists specific examples. Hypoxemia may be the result of impaired respiratory effort, obstruction of the upper airway, aspiration of blood or gastric contents, mechanical disruption of the chest cavity, or severe hemorrhagic shock. Traumatic brain injury (TBI) and intoxication with alcohol or other drugs contribute to impaired effort, upper airway obstruction, and aspiration, whereas direct trauma to the face, neck, or chest may cause bleeding, anatomic disruption of the airways or lung tissue, pneumothorax, or severe pulmonary contusions. Ventilatory failure is common in trauma patients, both at initial presentation and in the following days. Pulmonary contusion, with subsequent consolidation of alveolar space, may take hours to develop and may not be obvious until after fluid resuscitation and initial surgeries have been completed. Ventilatory failure may also result from exacerbation of underlying chronic cardiac or pulmonary disease or from other acute causes, such as pulmonary embolus (PE). Trauma patients are at very high risk for PE from their hypercoagulable state, vascular trauma, or fat emboli, and PE should be suspected in any patient with an abrupt decline in respiratory status. Patients with multiple injuries are at increased risk of developing the systemic inflammatory response syndrome (SIRS), which can be complicated by progressive respiratory compromise and recurrent sepsis and may lead to multiorgan system failure.

All trauma patients must be assumed to have a full stomach; obtaining an accurate history in the injured patient is difficult, and trauma itself will lead to a drastic decrease of gastrointestinal (GI) motility, with ileus persisting for hours to days after injury. Trauma patients are also at risk for aspiration of blood from open fractures or penetrating wounds of the face and airway. Impaired mental status resulting from TBI or intoxication may make aspiration more likely, particularly when combined with the use of sedative or analgesic drugs given to facilitate diagnostic procedures such as computed tomography (CT) or minor surgical procedures such as reducing a fracture or suturing a laceration.

Evaluation

Ideally, assessment of the patient before airway management is no different than assessment of an elective surgery patient. However, it must often be adjusted for the urgency of the situation. A thorough history and physical examination of the face, neck, and chest is appropriate when possible. Any suggestion that intubation will be difficult warrants the need for additional equipment or personnel. Presence of a cervical collar, facial fractures, or blood or vomitus in the airway add to traditional predictors of difficult intubation. When the urgency of the situation does not allow for a thorough assessment, the anesthesiologist must gather what information is immediately available from other providers, make a quick assessment of the patient, and then proceed as necessary. Box 17-2 summarizes factors predicting a difficult airway.

The need for “discretionary” intubation in the combative or uncooperative patient is controversial, and the provider must carefully assess the risks and benefits of intervention. Induction of anesthesia will allow for immediate diagnostic studies and more rapid identification of life-threatening conditions, such as epidural hematoma or splenic rupture. Induction and intubation may also prevent the patient from injuring self or others and may allow for deeper, safer levels of sedation during diagnostic studies. Induction, laryngoscopy, and intubation are not without risks. However, hemodynamic instability may be precipitated even in previously normotensive patients with either the induction agent or the institution of positive-pressure ventilation (PPV) with a subsequent decrease in venous return and cardiac output. Also, technical complications of rapid-sequence intubation (aspiration, oral trauma, need for surgical airway) may exacerbate the care of an otherwise minimally injured patient.

Early intubation, diagnostic imaging, and rapid extubation of the intoxicated patient without significant trauma are possible in some settings but can carry a substantial economic burden. Ultimately, the trauma team, including the anesthesiologist, must evaluate the potential for life-threatening trauma, the patient's ability to tolerate CT (with or without additional sedation), and the likely ease of intubation when deciding how to proceed. No matter what course is elected, close monitoring of the patient's neurologic status and respiratory effort is required .

Preparation

Sufficient trained personnel must be available to manage the airway physically, administer induction drugs, provide cricoid pressure (now controversial), and stabilize the cervical spine. The anesthesiologist performing or directing the intubation coordinates this process to ensure that all participants know their role. When possible, the plan of care should be discussed with the patient and family and any questions answered.

While other preparations are being made, preoxygenation should be maximized to the extent possible, whether through use of blow-by oxygen (O ₂ ), assisted bag-valve-mask (BVM) support of spontaneous ventilations, or ideally, a tight-fitting face mask. Although an apneic patient often must be preoxygenated through BVM ventilation, inspiratory pressures should be kept as low as possible to minimize the chance of gastric insufflation, regurgitation, and pulmonary aspiration of stomach contents. A high-flow suction device should be immediately available should regurgitation occur. All necessary intubating equipment, including primary and backup airway equipment, reliable sources of O ₂ and PPV, induction and emergency medications, and confirmatory equipment, should be close at hand.

Patient positioning can greatly facilitate intubation and is often overlooked in the emergent situation. The bed or stretcher should be placed at a convenient height for the anesthesiologist, with enough space at the head of the bed to allow room for unhindered motion. Ergonomic design of the trauma bay has been shown to improve the process of emergency intubation. Cervical spine instability is a consideration in most trauma victims; cervical injuries occur in 1.5% to 3% of all major trauma cases, and up to 50% of cervical fractures may be unstable.

Exclusion of cervical spine instability requires at least a cooperative patient without distracting injuries and may further require appropriate diagnostic studies (see later). The traditional “sniffing position” (head extension plus neck flexion) is thus contraindicated, whereas the presence of a rigid cervical collar and the maintenance of in-line cervical stabilization also contribute to the difficulty of intubation. Blood and other debris in the oropharynx can also make fiberoptic laryngoscopy difficult. If properly performed, direct laryngoscopy with manual in-line stabilization is unlikely to aggravate an existing cervical spine injury and has been judged safe and appropriate for the majority of trauma patients. Unfortunately, any manipulation of the airway, including mask ventilation, intubation, even placement of a laryngeal mask airway (LMA), can cause cervical spine motion. Cadaveric models of cervical injury demonstrated significantly less movement with in-line stabilization than with a cervical collar, and this technique has been judged safe and appropriate for the majority of trauma patients. Whereas some advocate routine use of fiberoptic intubation for all trauma patients with the potential for cervical instability, this approach is time and resource intensive, with no data showing improved outcomes. Multiple initial options that are less likely to exacerbate cervical instability (e.g., intubating LMA, light wand) may also be considered, given availability and provider experience.

Preprocedural preparation should include the availability of a device to facilitate intubation of an anterior larynx (e.g., Parker Directional Stylet, “trigger tube,” gum elastic bougie), rescue devices for failed intubation (e.g., LMA, Combitube, Cobra Perilaryngeal Airway, King Laryngeal Tube), and an interdisciplinary understanding of when cervical protection should be abandoned in favor of achieving a successful intubation. The likelihood of an anterior larynx argues for the routine use of a stylet in the endotracheal tube (ETT). A stethoscope and capnometry should be available to confirm endotracheal placement and adequacy of ventilation. Equipment should also be on hand for emergent percutaneous or surgical cricothyroidotomy in the worst-case scenario.

The American Society of Anesthesiologists (ASA) Difficult Airway Management Algorithm has been adapted for trauma, because “awakening” the patient who is hemorrhaging or unable to maintain an airway is not appropriate. In a review of 10 years' experience with intubation on arrival to a busy Level I trauma center, 6088 patients required intubation within the first hour of care. All were supervised or done by attending trauma anesthesiologists. Of these patients, 21 (0.3%) received a surgical airway. Unanticipated difficult upper airway anatomy was the leading reason for surgical airway. All these intubations were performed or attempted with direct laryngoscopy. The leading causes of the need for surgical airway were difficult anatomy (11), foreign body (6), and injury to head or neck (5).

Several small studies have investigated various types of indirect/video laryngoscopes (GlideScope, Bullard, McGrath, Airtraq, Pentax Airwayscope, Truview EVO ₂ , Viewmax), which provide a theoretic advantage in minimizing cervical spine motion during intubation. Simulation, cadaver model, and live-patient evaluations generally show that the Cormack-Lehane grade is improved, cervical spine motion is reduced, and time to intubation is similar with indirect laryngoscopy compared with direct laryngoscopy (DL). However, a cinefluoroscopic study of 20 patients without cervical pathology, using manual in-line stabilization by an assistant, showed no decrease in movement of the cervical spine with GlideScope versus DL. Recently, prehospital difficult intubation management by experienced European anesthetists improved with the GlideScope.

Induction/Intubation Considerations

A rapid-sequence intubation (RSI) technique is recommended, with the use of cricoid pressure (CP, Sellick maneuver) from induction of anesthesia (or onset of apnea) until confirmation of correct ETT positioning. Although how consistently CP prevents regurgitation and aspiration of gastric contents has been questioned, the technique is also beneficial in moving the larynx into a position of better visualization, the backward-upward-rightward pressure (BURP) technique, thus maximizing the laryngoscopic view of the vocal cords. If active vomiting (vs. passive regurgitation) begins while CP is being held, the cricoid cartilage should be released to minimize the risk of spontaneous esophageal rupture (Boerhaave's syndrome). Suction and positioning should be employed (e.g., turning patient en masse if on spine board) to minimize the risk of pulmonary aspiration.

With adequate dosing of induction/intubation agents and suction immediately available to the intubator, aspiration is unlikely during direct airway visualization. Specific manipulation techniques may optimize visualization of the glottic opening, whereas other techniques may worsen the view. A study of 104 emergency medicine physicians performing 1530 sets of laryngoscopy on fresh cadavers suggested that the percentage of glottic opening using a validated scoring scale improved more with bimanual laryngoscopy than with CP, BURP, or no manipulation.

Although a standard-of-care in emergency intubations, the efficacy of cricoid pressure has been questioned. In 1961, Sellick described CP as a method to reduce the risk of aspiration during the induction phase of anesthesia, and although widely used, its method of application, timing, and role in difficult airways are not standardized. For emergency airway management, the risk of aspiration is thought to be higher than in elective cases, ranging up to 22% in ED-performed RSI, and CP is a theoretic preventive maneuver. However, the amount of pressure needed and the optimal method of application are unknown, and the pressure may displace rather than occlude the esophagus. Also, CP may make both mask ventilation and laryngeal view more difficult, both of which can be improved by release of CP. In the United Kingdom, where anesthetists work as prehospital physicians, a prospective study of CP for RSI in the field reported that of 402 intubations, CP was released in 22 patients to improve laryngeal view, bimanual manipulation was used in 25 intubations, and BURP was applied in 14 intubations; 98.8% of patients were intubated on the first or second attempt. Two patients, who had prolonged bag-mask ventilation and difficult intubations, regurgitated after release of CP. This illustrates a larger unanswered question: are trauma patients at risk for aspiration simply because they have a full stomach, or is it caused by often-inadequate induction doses of medications (from hemodynamic instability) and challenging airways with cervical spine immobilization, blood/vomitus, or facial injuries? Many suggest using CP for RSI, with release of CP if the mask ventilation or visualization becomes difficult.

Induction/Intubation Medications

Advantages and disadvantages of various induction drugs are shown in Table 17-2 . Although agents that lack a negative inotropic effect (e.g., ketamine, etomidate) are more likely to preserve cardiovascular function in the euvolemic patient, any induction drug—and even the change to PPV alone—can precipitate hemodynamic instability in the patient in shock. This is because the hypovolemic patient is relying on a high serum level of catecholamines to support the blood pressure. Some degree of catecholamine depletion should be assumed in the trauma patient. Many sedative or analgesic agents may depress sympathetic tone, impair the adrenal response to hemorrhage, “unmask” hypovolemia, and cause profound hypotension. Internal hemorrhage may not be readily apparent at induction, and vital signs are at best a crude indicator of volume status; therefore care should be taken with any anesthetic agent. Such situations demand the use of smaller-than-normal doses, carefully titrated to the patient's response.

Although etomidate may seem to be the ideal agent for use in trauma patients because it maintains hemodynamic stability, reported complications related to etomidate induction in trauma patients may preclude its use. Occult adrenal insufficiency has been noted in up to 60% of severely injured patients and is associated with persistent systemic inflammation, a hyperdynamic cardiovascular state, and vasopressor-dependent shock. In a retrospective study of ICU patients at a Level I trauma center, 137 patients had undergone cosyntropin stimulation testing; there was no difference in age, gender, race, injury severity or mechanism, rates of sepsis/septic shock, mechanical ventilation, or mortality. Patients who had received etomidate were more likely to have adrenal insufficiency, as defined by “nonresponders” to cosyntropin. A more recent study analyzed 94 patients who had received etomidate for prehospital intubation. Again, with no differences between those who did or did not receive etomidate, its use was associated with a higher incidence of acute respiratory distress syndrome (ARDS) and multiple-organ failure, thought to be caused by etomidate's effect on the inflammatory system (inhibition of 11β-hydroxylase). A larger, randomized prospective trial of etomidate (234 patients) versus ketamine (235) for RSI did not assess mortality; the primary endpoint was the maximum score of the sequential organ failure assessment (SOFA) during the first 3 days in the ICU. The mean maximum SOFA score between the two groups did not differ significantly, but the percentage of patients with adrenal insufficiency was significantly higher in the etomidate than in the ketamine group.

Succinylcholine is the standard paralytic agent for RSI and is recommended in the absence of obvious contraindications (pre-existing neuromuscular disease, known or suspected hyperkalemia; burn, spinal cord deficit, or massive muscle trauma occurring more than 24 hours previously; recent prolonged bed-bound status; known history of malignant hyperthermia). Rocuronium or vecuronium can be used in place of succinylcholine and will provide similar intubating conditions and almost the same speed of onset, at the cost of greatly prolonged paralysis thereafter.

The administration of positive-pressure breaths by BVM during RSI is controversial. In trauma cases in which RSI is undertaken in a reasonably cooperative, maximally preoxygenated patient, PPV can often be completely avoided, to minimize gastric distention and increased likelihood of aspiration. In the emergent, desaturating patient, or when preoxygenation is limited or impossible, BVM ventilation throughout RSI may be considered. No data support or refute this practice, and the clinician must use best judgment in obtaining an airway. Preoxygenation may be difficult if the patient is combative or if anatomic positioning is suboptimal, and even transient hypoxemia may be dangerous to the patient with TBI or hemorrhagic shock.

With trained providers, RSI of the trauma patient has been reported to be successful on the first attempt more than 90% of the time. In the remaining cases, knowledge of the local difficult airway algorithm becomes essential. Providers vary in their skills, institutions vary in available equipment, and time pressures of an emergent intubation makes creative thought difficult. The adage that “no one gets smarter in an emergency” is particularly apropos in dealing with the airway of a trauma patient. It is therefore incumbent on every anesthesiologist to plan for the steps to follow if a given intubation proves challenging. Some “difficult airway” carts include complex equipment, which is associated with complications, and alternative devices are used less frequently. Small sets of more frequently used equipment are more helpful in emergency situations.

Every anesthesiologist should be familiar with the ASA Difficult Airway Management Algorithm, modified for trauma, which should be followed in most cases. The algorithm for emergent intubations is considerably simpler, because awakening the patient usually is not a viable option.

Successful intubation by whatever route should ideally be confirmed by multiple methods, including the detection of carbon dioxide (CO ₂ ) in exhaled breaths. In areas where intubation and mechanical ventilation are common, such as the ED trauma bay or ICU, continuous-waveform capnography is highly recommended. For other areas, a disposable CO ₂ capnometer should be part of the emergency intubation setup. Exhalation of CO ₂ is possible only in patients with a perfusing rhythm; thus, patients with no cardiac output may produce no exhaled CO ₂ . The lack of positive capnometry despite a properly placed ETT may be the first indication in the field or remote setting that cardiac arrest has occurred. Even with cardiopulmonary resuscitation in progress, the patient may still produce little or no detectable CO ₂ . In these patients, successful intubation should have been confirmed by initial DL examination, with auscultation confirming bilateral breath sounds, the absence of gastric sounds, the presence of equal chest rise, and misting in the tube, or by the use of an esophageal detector device (EDD). Inspection by repeat DL may also be considered, as opposed to improperly pulling a correctly placed ETT.

After confirmation of successful intubation, the anesthesiologist in some trauma centers is responsible for assessment of hemodynamic stability after induction, initial ventilator settings, vascular access, and ongoing sedation and analgesia. Patient awareness during intubation and mechanical ventilation is a significant problem in trauma cases, particularly when hypotension limits the amount of induction or sedation agent, and paralytic agents have been used to facilitate diagnostic studies or minor procedures. Ketamine, scopolamine, or small amounts of a benzodiazepine may be considered in patients at particular risk for awareness. Awareness monitors, such as the Bi-Spectral Index, may be considered in such cases, although at present these do not constitute the standard of care and should not interfere with the application of definitive care that will allow adequate levels of sedation. Even if not directly involved in this phase of care, the anesthesiologist can contribute substantially to the prevention and recognition of this problem, as well as to the education of other hospital staff.

Airway and breathing are the first priorities in trauma care, followed closely by assessment of the circulation—the ABCs. The anesthesiologist may share responsibility for hemodynamic management in the ED with other members of the trauma team, but in the OR this becomes their primary task. In the ED the anesthesiologist should be ready to take primary responsibility for the ABCs if this critical task is being neglected.

Pathophysiology

Trauma causes disruption of blood vessels of all sizes, and hemorrhage, whether frank and focused or insidious and diffuse, is a hallmark of trauma. Although low-pressure bleeding can be managed expectantly or with simple techniques, in other trauma patients, active intervention is required to prevent hypovolemia, hypotension, and exsanguination. Although control of hemorrhage is paramount and may be easy to achieve in some injuries, airway management must remain the priority, and practitioners need to avoid distraction in completing the ABCs ( Fig. 17-2 ). Uncontrolled, life-threatening and noncompressible hemorrhage can occur from venous bleeding in “open book” pelvic fractures and in some severe liver injuries ( Figs. 17-3 and 17-4 ). Life-threatening hemorrhage occurs into one of five “compartments,” summarized in Table 17-3 . Although not a complete list, patients with any of the injuries in Box 17-3 should be considered at high risk of death. Early diagnosis and control of hemorrhage are essential, but equally important are nonsurgical hemorrhage control and ongoing resuscitation, which may be underappreciated.

Shock is the term used to describe the complex pathophysiology that arises from inadequate tissue perfusion ( Box 17-4 ). Shock was first described in trauma patients, in whom hemorrhage is a common cause. Trauma patients may exhibit shock caused by an aberration in any or all components of cardiovascular physiology: preload (mechanical impairment of blood flow, as in tension pneumothorax or cardiac tamponade), contractility (cardiac dysfunction, as in blunt myocardial injury or secondary to severe TBI, ingestion of toxins, or anesthetic overdose), and afterload (spinal cord injury, medications). Affecting preload, hemorrhage is considered to be the source of shock in all trauma patients until it is definitively ruled out. Much of the ATLS curriculum is devoted to this important diagnostic and therapeutic process.

Hemorrhage reduces circulating blood volume, leading to decreased preload and reduced cardiac output. Vasoconstriction and increased inotropy mediated by the sympathetic nervous system allow for continued blood flow to vital organs in the presence of blood loss as severe as 40 mL/kg, the cutoff between a class III and class IV hemorrhage per ATLS guidelines, or about 2 L of the 5 L of total blood volume in a 70-kg previously healthy patient. Acute blood loss in excess of this amount causes a critical reduction of perfusion to the heart and brain, manifesting as coma, pulseless electrical activity, and death. Lesser blood loss may also be lethal, particularly in elderly patients or those with medical comorbidities, because reduced perfusion leads to anaerobic metabolism and accumulation of lactic acid and other toxins. Individual cells react to ischemia by hibernation (reduction of all nonessential activities), apoptosis (programmed cell death), or outright necrosis, depending on the organ system in question. Many ischemic cells, especially gut and muscle cells, react to ischemia by absorption of extracellular fluid ; this loss of potential circulating volume can be exacerbated by overaggressive or repetitive “running of the bowel” by the surgeon, which also causes edema and dysfunction of the luminal wall. The end result of this tissue edema is both locally and systemically disruptive, by clogging capillary pathways (no-reflow phenomenon) and further hampering autoresuscitation (reclamation of interstitial volume into vascular spaces). Ischemic cells also release inflammatory mediators, triggering a chemical cascade that perpetuates the pathophysiology of shock long after adequate circulation is restored.

The “dose” of shock absorbed by the body, a summation of the depth of hypoperfusion and its duration, largely determines the patient's clinical outcome, ranging from a mild inflammatory response to organ system failure to death. The typical young male trauma patient has an enormous compensatory reserve and may achieve normal pulse and blood pressure while still significantly fluid depleted and highly vasoconstricted. This phenomenon, known as the occult hypoperfusion syndrome, is associated with a high incidence of organ system failure if not recognized and corrected.

Isotonic crystalloid infusion increases circulating volume and preload, producing an immediate increase in cardiac output and blood pressure. Volume therapy is a double-edged sword, however, as increased BP before adequate surgical and medical hemostasis is achieved can lead to increased bleeding from open vessels and rebleeding from previously hemostatic injuries (“popping the clot”), in part caused by decreased blood viscosity and relaxation of compensatory vasonstriction. Further, aggressive crystalloid infusion dilutes red blood cell (RBC) mass and clotting factor concentration, possibly resulting in decreased O ₂ delivery at the tissue level and an increase in blood loss, respectively. Unless properly heated, such infusions may also lead to hypothermia, which should be cautiously guarded against using fluid warmers and forced-air warmers. Studies of uncontrolled hemorrhagic shock in rats, swine, sheep, and dogs have all demonstrated improved survival when initial fluid therapy is titrated to a lower-than-normal systolic BP (70-80 mm Hg). This finding is supported by two human trials.

Dilution of RBC mass is inevitable during early resuscitation, because losses to hemorrhage are compounded by intravascular recruitment of extracellular fluid and exogenous crystalloid administration, the phenomenon of autotransfusion. A hematocrit measured soon after hemorrhagic trauma may show little change, as whole blood is being lost and the RBC percentage in the remaining volume does not change. Thus, a stable hematocrit in the face of ongoing loss is meaningless information. The longer hemorrhage and resuscitation persist, the more the hematocrit will fall. Loss of RBCs leads to decreased blood viscosity, allowing for a compensatory increase in blood flow but a decrease in O ₂ content, and tissue O ₂ delivery begins to decrease. Fluid resuscitation after massive hemorrhage will result in extensive hemodilution and coagulopathy; this hemodilution affects procoagulants as well as anticoagulant, profibrinolytic, and antifibrinolytic components of the coagulation cascade. Factor replacement early in resuscitation, with plasma, platelets, and occasionally cryoprecipitate, may mitigate a severe coagulopathy.

Evaluation

The diagnostic characteristics of hemorrhagic shock and goals for resuscitation are listed in Box 17-4 and Table 17-4 . Control of bleeding is the overarching priority in treatment, and nothing must interfere with the indicated diagnostic or therapeutic procedures shown in Table 17-3 . Relevant patient physiology is assessed by continuous measurement of vital signs (facilitated by early placement of arterial pressure catheter) and by immediate and appropriately repeated measurements of arterial blood gases (ABGs), complete blood chemistry, clotting function, and serum lactate determination. Toxicology screening and electrocardiography may be useful in discovering and addressing underlying reasons for suboptimal response to resuscitation. Patients who arrive to the hospital with signs of coagulopathy on admission (elevated activated partial thromboplastin time) have likely developed the “acute coagulopathy of trauma” and are at increased mortality risk.

Response to fluid therapy will provide important diagnostic information. Most patients in shock will demonstrate an improvement in vital signs after bolus fluid administration. In “responders,” those who have achieved spontaneous hemostasis (i.e., those with lung injury or peripheral orthopedic injuries), the improvement in vital signs will be sustained. “Transient responders,” those with ongoing hemorrhage (e.g., abdominal visceral trauma, pelvic fracture) will show initial improvement in vital signs that decays over about a half hour and are in need of urgent diagnostic studies and therapeutic procedures. “Nonresponders,” those who do not respond to an initial fluid bolus, either have a nonhemorrhagic source of shock (e.g., obstruction to venous return to heart, spinal cord injury, cardiac disease) or are bleeding rapidly.

The initial choice of crystalloid likely does not matter; it is critical, however, to understand the risks and benefits of each type of fluid infused and, in life-threatening hemorrhage, to begin blood transfusion early. Hemorrhage control depends on 13 factors in the coagulation cascade, none of which is contained in crystalloid, packed red blood cells (PRBCs) or in cell-saver blood. Dilutional resuscitation of hemorrhagic shock with colloid (e.g., hetastarch) or crystalloid reduces the concentration of coagulation factors in the circulating blood volume and impairs hemostasis. Resuscitation with normal saline results in hyperchloremic acidosis that may be associated with systemic vasodilation, increased extravascular lung water, and coagulopathy. Ringer's lactate became the standard of care for fluid resuscitation in the 1960s in an effort to replete bicarbonate in patients with severe dehydration. Ringer's lactate has recently been found to be proinflammatory and to activate neutrophils, which are primary effector cells in reperfusion injury, particularly in the formulations that contain the d -lactate isomer (it is not oxidized by l -lactate dehydrogenase, and therefore accumulates, and activates neutrophils). Plasmalyte is a “balanced physiologic” solution that contains sodium, potassium, chloride, acetate, gluconate, and magnesium, but not calcium, and is therefore compatible with blood infusions. Small studies in animals have shown increased mortality with Plasmalyte resuscitation, possibly caused by the peripheral vascular resistance (PVR) effects of magnesium ; no large resuscitation studies have been done.

Although there is no “optimal” crystalloid, these solutions are still the fluids of choice for initial resuscitation in most patients following hemorrhage. The anesthesiologist must understand, however, that large volumes of replacement with crystalloid result in distribution throughout the entire extracellular compartment, resulting in massive fluid overload and edema, with complications such as acute respiratory failure, hepatic failure, renal failure, sepsis, and most recently, abdominal compartment syndrome. Despite multiple randomized controlled trials (RCTs) comparing albumin and colloids to crystalloids, no strong data show that colloids are associated with improved survival in trauma or burn patients. Multiple studies of hypertonic saline solutions (3%, 5%, 7.5%) for trauma resuscitation have been encouraging, although an RCT of 7.5% saline, 7.5% saline/6% dextran, and 0.9% saline (Resuscitation Outcomes Consortium) failed to show a difference in 28-day survival.

Because the civilian community has not yet adopted a limited-fluid resuscitation strategy as has the military, and because clinical trials have shown no significant benefit with hypertonic fluids, standard use of these fluids cannot be recommended at this time. The recognition that use of fluid replacement in severe injuries that was as near to whole blood as possible (i.e., blood, plasma, and platelets in a 1:1:1 ratio), resulted in increased survival and decreased complications. This concept of “damage control resuscitation” has recently become standard of care.

Preparation

Resuscitation of the actively hemorrhaging patient requires large-bore, high-flow intravenous access, preferably through at least two separate catheters. Warmed IV fluids are highly recommended, especially early in resuscitation. Commercial fluid-warming technology is highly effective and should be used as frequently (or more so) in the trauma bay as in the OR. Rapid infusion systems can warm and deliver large volumes quickly and may be lifesaving in the patient with rapid and uncontrolled hemorrhage. However, resuscitation with a rapid-infusion system targeted to a normal blood pressure, prior to hemorrhage control in trauma patients, is associated with an increase in mortality.

The ability to rapidly administer uncrossmatched type O blood may be lifesaving. Many trauma center blood banks and EDs maintain a supply for this purpose. Crossmatched blood, plasma, and platelets should be requested at the earliest moment that a massive transfusion is anticipated. Additional personnel, from both the anesthesia and the nursing staff, should be mobilized early to address the multiple needs of emergency surgery and resuscitation.

High-level trauma centers will maintain a designated trauma OR that is kept warmed and ready with drugs, IV fluids, and rapid-infusion devices. As opposed to elective operative cases where the surgery begins after anesthesia has instituted appropriate access and monitoring, the primary goal in treating a patient with exsanguination is hemorrhage control; therefore it may be necessary for anesthesia access and monitoring to be obtained concurrently with surgical intervention.

Intraoperative Considerations

Resuscitation must be carefully choreographed with diagnostic and therapeutic procedures such that tissue perfusion is optimized without unnecessarily large increases in blood pressure that can exacerbate uncontrolled hemorrhage. Recent understanding of the potential for rebleeding and dilution has led to a change away from the traditional ATLS approach of rapid crystalloid infusion to one of deliberate, controlled fluid administration, titrated to specific physiologic end points (see Box 17-4 ).

Replacement of RBCs is essential to limiting the severity and duration of shock after hemorrhage. PRBCs should be administered early in the resuscitative process, using uncrossmatched type O units if necessary. Adverse reaction to this therapy is extremely unlikely: more than 100,000 units of uncrossmatched blood were administered during the Vietnam War without a single documented case of fatal transfusion reaction, compared with the nine cases that occurred in the 600,000 crossmatched transfusions. Immediate transfusion of type O blood is sufficiently safe and beneficial that it should be considered for any patient presenting in extremis from hemorrhagic shock. The most appropriate target hematocrit for resuscitation must be individualized on the basis of age, specific injury pattern, pre-existing disease, and the potential for further hemorrhage.

Coagulopathy resulting from acute consumption of coagulation factors is likely in any patient losing more than a single blood volume (~ 5 L in 70-kg adult) or receiving more than 10 units of RBCs ( Fig. 17-5 ). Because coagulopathy is more easily prevented than treated, early administration of plasma to any patient who has lost or will lose this amount of blood is highly recommended. Timely initiation of a massive transfusion protocol is associated with improved survival and reduced transfusions. This was first described in military resuscitation, but it has since also become a standard of care at civilian trauma centers. This has come to be known as a “1:1:1” resuscitation, suggesting that 1 unit of plasma should be transfused per each RBC unit, and that platelets should be administered similarly (remembering that a “pack” of platelets is pooled from multiple donors and may represent 4-6 “units”).

Early activation of a massive transfusion protocol is associated with improved patient survival. Plasma should be ordered from the blood bank for any patient presenting emergently to the OR with symptoms of acute hemorrhagic shock. A ratio of 1:1 replacement of RBCs and plasma is appropriate for any patient who has lost or is likely to lose more than 1 blood volume, although this should be guided by clinical assessment and, time permitting, judicious laboratory evaluation. Although concurrent plasma factor replacement with ongoing blood infusion will promote clot formation, factor levels (V, VII, VIII, protein S, and von Willebrand) decrease with storage, so each patient's response will be variable and may depend on the age of the blood products. Platelet count usually remains adequate longer than coagulation factors, and platelet therapy is therefore required less often than plasma. Transfused platelets have a very short functional life span in the circulation and pose both a significant immune stimulus and an infective risk. However, factor replacement and platelet therapy should not be delayed while awaiting laboratory values in patients with exsanguinating hemorrhage. A multicenter validation of a score to predict which patients will need a massive transfusion suggests that penetrating trauma, ED systolic BP less than 90 mm Hg, heart rate greater than 120 beats/min, and positive FAST score (focused abdominal sonography for trauma) are indicators of severe hemorrhage.

Coagulation factor concentrates and cryoprecipitate may not offer a benefit beyond that of plasma infusion in the hemorrhaging trauma patient, unless fluid overload is a significant risk, as in the coagulopathic elderly patient, or the patient is known to have a specific factor deficiency. In exsanguinating patients, however, fibrinogen replacement (with cryoprecipitate) may enhance clot stability, because fibrinogen is the first coagulation factor to become critically low in patients with major hemorrhage. Anecdotal reports after the early use of activated recombinant factor VII (rFVIIa) describe rapid resolution of traumatic coagulopathy after administration of 20 to 100 units. However, the large CONTROL trial, which randomized patients who had bled 4 to 8 RBC units to rFVIIa or to placebo, did not show a difference in 30-day mortality. Patients who received rFVIIa had fewer units of transfused blood overall, and with no increase in thrombotic complications compared with placebo, it seems safe to use when indicated.

Electrolyte abnormalities are common during resuscitation from hemorrhage. Hyperosmolarity may result from alcohol ingestion, dehydration, hypovolemia, or administration of normal saline (NS). Mild hyperglycemia secondary to high circulating catecholamine levels is expected. Neither of these conditions mandates specific treatment during resuscitation, because both will resolve with restoration of adequate intravascular volume. Hyperchloremic metabolic acidosis is a significant risk for overresuscitation, especially with mildly hypertonic solutions such as NS, and can be managed with the titrated addition of hypotonic fluids.

Hypocalcemia arises from chelation of circulating calcium by the citrate or adenosine additives found in banked blood products. IV administration of calcium is indicated in patients with low serum Ca ⁺⁺ levels and should be considered for empiric administration in the case of massive transfusion, particularly in the presence of hemodynamic instability. Serum bicarbonate levels will be lower than normal in the hemorrhaging patient as a result of increased lactic acidosis and impaired renal blood flow. Some recommend administration of bicarbonate solutions to increase systemic pH in acidotic patients, to enhance the functioning of important protein systems, including coagulation and catecholamine receptors. Bicarbonate also supports cardiac contractility and can be useful in cardiopulmonary collapse. The clinical utility of bicarbonate therapy, however, has never been proved. Vasopressin (antidiuretic hormone) is emerging as an important advance in the treatment of a variety of shock states. Plasma levels of vasopressin increase within a few minutes of circulatory arrest and also rise in response to hemorrhage, in which endogenous vasopressin release is an important vasoconstrictor mechanism. Vasopressin does not depend on pH for its activity, so it would therefore offer a theoretic advantage over other vasoactive agents, but there are no large, prospective human trials. Adequate fluid resuscitation remains the primary therapy for restoration of normal acid-base status.

Early resuscitation has evolved toward less aggressive fluid administration. Late resuscitation is characterized by the need to completely restore and support perfusion, usually in the ICU. To do so requires the practitioner to look beyond the vital signs for a more direct measure of tissue perfusion. The speed with which serum lactate level normalizes after shock is strongly associated with the risk of death from organ system failure. End points of resuscitation should focus on organ-specific signs of recovery of function: improved lung function, cardiac contractility, and vasomotor tone; clearance of toxins by the liver and kidney; and absence of infectious complications, common after severe traumatic injury. Although the overall mortality from multiple trauma has declined in the past decade, there has been no significant decrease in mortality from sepsis after severe trauma. Risk factors for posttraumatic sepsis are male gender, age, pre-existing medical conditions, Glasgow Coma Scale (GCS) score of 8 or less, high injury severity score (ISS), number of injuries, number of RBC units transfused, number of surgical procedures, and laparotomy.

Pathophysiology

Traumatic brain injury is classified as mild, moderate, or severe, depending on the GCS score on admission. Mild TBI (GCS score, 13-15) is the most common. Although mild TBI does not usually necessitate intensive treatment, patients may be significantly debilitated by postconcussive symptoms, including headaches, sleep and memory disturbances, and mood swings. Progression of mild TBI is rare but may be catastrophic. Research, case reports, and media coverage of U.S. soldiers returning from the Iraq and Afghanistan wars have highlighted the long-term and sometimes violent aftereffects of mild TBI that were previously underrecognized: depression, mood changes, aggressive behavior, depression, and memory loss.

The most frequently proposed cellular mechanism in mild TBI is diffuse axonal injury (DAI), associated with alterations in many physiologic processes. There is an alteration in proteostasis; proteopathy is often evident at the histopathologic level. Here, the pathways of idiopathic and posttraumatic neurodegeneration apparently overlap, since identical protein aggregates accumulate in both conditions. As early as 2 hours after severe TBI, increased levels of soluble amyloid-β (Aβ) peptide and deposition of amyloid plaques are evident in brains of 30% of survivors, regardless of their age. An acute, single-incident TBI is also found in the history of 20% to 30% of patients with Alzheimer's disease or parkinsonism, but in only 8% to 10% of control subjects.

A broader military probe in 2010 found that up to half of soldiers with posttraumatic stress disorder (PTSD) or depression after mild TBI while deployed reported misuse of alcohol or aggressive behaviors (punching, fighting) following their return to society. Rates of depression and PTSD were higher at 12 months than at 3 months after deployment. Many soldiers do not seek help, but a new U.S. Army program Re-Engineering Systems of Primary Care Treatment in the Military (RESPECT-MIL) for returning soldiers, families, support systems, and the public seeks to decrease the stigma traditionally associated with difficulties with reintegration into society.

The U.S. National Football League, along with college and high school football associations, has also begun a more cautious approach to the management of players with a “concussion.” Chronic traumatic encephalopathy (CTE), a condition typically seen in retired or aging athletes, was recently reported in a college football player who committed suicide. The diagnosis was unique in two ways: he was the first active college athlete to be found with the disease, and unlike other known victims of CTE, he was never diagnosed with a concussion. However, this player had been a lineman and a linebacker, positions that typically involve multiple hits to the head during every game and practice, with estimates of approximately 1000 hits per season or more. CTE is traditionally associated with neurobehavioral disorders and bizarre behavior and is also called dementia pugilistica, or boxer's dementia. Career boxers sustaining repeated blows to the head and concussions may develop the syndrome. CTE is likely caused by large accumulations of tau proteins in the brain that kill cells in the regions responsible for mood, emotion, and executive functioning. Tau proteins are also found in the brains of patients with Alzheimer's disease and dementia.

Players with mild TBI may be evaluated in the trauma bay and discharged to home, with no need for anesthesia contact. However, recognizing the potential long-term consequences can be helpful if they present to the OR for other cases, such as fracture fixation. A systematic review of “brain concussion” management identified 4319 articles; when the search term “mild TBI” was used, 2509 articles were identified, and this decreased to 39 articles with “return to play” as keywords. Although only few studies address this topic, the Vienna Statement, Prague Statement, American Academy of Neurology, U.S. Team Physician Consensus Statement, and U.S. National Athletic Trainers Association Position Statement all agree on the following points:

• ▪

  There should be a period of rest, aerobic exercise, and drills before players with mild TBI return to play (each ~ 24 hours), in addition to evidence of normal cognitive function, and no recurrence of symptoms with exertion.

• ▪

  There is an age-dependent difference in recovery of function; highschool athletes take an average of 30 days to recover normal cognitive function, college athletes 7 to 10 days, and professional athletes only 3 to 5 days.

Moderate TBI (GCS 9-12) is more likely to be associated with intracranial lesions that require surgical evacuation. These patients have a higher potential for deterioration and are more susceptible to secondary insult if not carefully managed.

Severe TBI (GCS ≤ 8) is a highly lethal condition, often associated with intraparenchymal or intraventricular hemorrhage or evidence of DAI on cranial CT. Magnetic resonance imaging (MRI) is more sensitive than CT in the detection of traumatic brain lesions, especially in nonhemorrhagic DAI. A patient who has a negative brain CT with a poor neurologic status should be assumed to have DAI, especially without confounding factors such as intoxication or drugs. Patients with severe TBI are usually unable to maintain airway patency and may have diminished or absent respiratory drive, with inability to protect the airway from aspiration. Most patients presenting to the OR will have severe TBI, with elevation of intracranial pressure (ICP) caused by hemorrhage (epidural, subdural, or intraparenchymal), edema, or both. Failure to promptly relieve elevated ICP will lead to herniation of brain tissue, loss of brain blood flow, and death. The surgical goal is resolving increased ICP and controlling any active hemorrhage. Even brief periods of hypotension or hypoxemia can affect outcomes in head injury. An investigation of the impact on outcome of hypotension (systolic BP < 90 mm Hg) and hypoxia (Pa o ₂ ≤ 60 mm Hg or apnea or cyanosis in the field) revealed that these were independently associated with significant increases in morbidity and mortality from severe head injury. Hypotension was profoundly detrimental, occurring in 34.6% of these patients and associated with a 150% increase in mortality.

Evaluation

Along with imaging studies, the neurologic examination is of critical importance in the preoperative assessment of the TBI patient. Recovery from TBI is a gradual process. The sedative effects of anesthetic medications may be exaggerated, and the trauma patient will seldom improve immediately at the conclusion of cranial decompression. It is important to monitor for deterioration in the neurologic examination so that critical serial imaging studies and appropriate ICU management can commence as soon as possible.

More controversial is the timing of noncranial surgery in the patient with TBI. Transient hypotension or hypoxemia associated with orthopedic surgery may lead to worsening of neurologic injury, whereas delay in repair of fractures may increase the risk of pulmonary complications and sepsis. Although no definitive prospective study has been conducted, recent retrospective work suggests that early surgery with strict attention to hemodynamic goals does not necessarily worsen TBI.

The anesthesia provider has the responsibility for ensuring adequate O ₂ delivery to the injured brain and the penumbra, in an effort to prevent any further damage. Current Brain Trauma Foundation guidelines recommend a cerebral perfusion pressure (CPP) of 50 to 70 mm Hg; targeting a higher mean arterial pressure (MAP) is associated with a higher incidence of ARDS and mortality.

Preoperative Preparation

Early intubation of the TBI patient may be required because of combative or agitated behavior, the need for diagnostic studies before reaching the OR, and the potentially catastrophic consequences of respiratory depression or pulmonary aspiration. In fact, most patients with moderate or severe TBI will present to the OR having been intubated in the field or ED. There is no consensus on whether patients with severe TBI should be intubated in the field or in the ED on hospital arrival; studies show improvement with each management strategy. Where intubation occurs likely depends more on the ability of prehospital providers to manage an airway acutely, and more importantly, their training in RSI and access to emergency airway drugs.

Arterial pressure monitoring is required for any intracranial procedure, because the dramatic BP swings that can occur throughout such cases need to be closely monitored and limited to the extent possible. Large-bore IV access is necessary because blood loss from the open scalp or from the brain parenchyma can become excessive, particularly in patients with severe TBI and early onset of coagulopathy. Other medications likely beneficial include induction or maintenance agents such as thiopental, antiepileptics such as phenytoin (Dilantin) or levetiracetam (Keppra), diuretics such as furosemide (Lasix) or mannitol, and hypertonic saline (HTS). Patients with severe TBI should receive a 7-day course of seizure prophylaxis with either phenytoin or levetiracetam. If not administered in the ED, the loading dose needs to be given in the OR. Although many clinicians still use mannitol as their osmotic diuretic of choice to decrease ICP, increasing evidence shows that HTS solutions are more effective. A meta-analysis of RCTs found that HTS is more effective than mannitol for the treatment of elevated ICP. HTS can also be effective in lowering ICP after failure of standard mannitol therapy. In addition to effects of volume expansion, improved cardiac output, improved cerebrospinal fluid (CSF) absorption, and immunomodulation, HTS may be superior to mannitol with respect to brain oxygenation and cerebral hemodynamics. It is always helpful to know what therapies a patient has received in the ED or ICU before coming to the OR.

Intraoperative Management

Patients with mild TBI pose few additional anesthetic risks but are more susceptible to the effects of sedative medication. Benzodiazepines should be used judiciously throughout the perioperative period because they can easily complicate the neurologic examination. The anesthesiologist should strive to have the patient's sensorium as clear as possible as rapidly as possible after any anesthetic. Any change from the patient's preoperative mental status not attributable to anesthetic drugs is an indication for immediate repeat head CT and neurosurgical reassessment.

The care of patients with moderate TBI consists of serial assessment of neurologic function, with repeat CT at regular intervals. If close monitoring is not possible, owing to the need for general anesthesia or sedating medications, then continuous invasive measurement of CPP is indicated. An ICP monitor is recommended in any patient with moderate or severe TBI undergoing noncranial surgery likely to last longer than 2 hours. Patients with severe TBI are particularly challenging. Early, rapid focus on restoring systemic homeostasis and maximizing perfusion to the injured brain will produce best possible outcomes. Again, hypoxemia (Pa o ₂ < 60 mm Hg) or hypotension (systolic BP < 90 mm Hg) in patients with severe TBI is associated with a significant increase in mortality. Management requires a highly skilled facility, close cooperation among providers, and a stepwise implementation of therapies, as shown in Figure 17-6 .

Aggressive restoration of intravascular volume is indicated to maintain intracranial perfusion, especially if associated pulmonary injuries necessitate the use of high mean airway pressures to support oxygenation. Hyperventilation, previously a mainstay in TBI management for its ability to lower ICP through reduction of intracranial blood flow, is no longer appropriate unless the patient shows signs of imminent brainstem herniation, because this reduction of flow puts ischemic brain tissue at further risk for necrosis or apoptosis. Hyperventilation is indicated only for patients who present with strong lateralizing signs en route to CT and emergent decompressive surgery.

A systolic BP of 90 mm Hg should be maintained in patients with severe TBI, with MAP of 50 to 70 mm Hg until invasive ICP monitoring can be placed. Previous guidelines suggested maintenance of CPP at a minimum of 70 mm Hg at all times; increasing MAP to greater than 70 mm Hg may not improve outcome, particularly in patients in whom autoregulation is lost. Contrary to past practice, the patient with severe TBI should be maintained in a euvolemic state. Fluid resuscitation is the mainstay of therapy, followed by vasoactive infusions as needed. Controversy surrounds the appropriate “transfusion trigger” in patients with severe TBI, and whether these patients should be treated as other critically ill patients in whom a hemoglobin (Hb) concentration of 7 g/dL is a proven trigger. Many neurosurgeons prefer an Hb level closer to 10 g/dL in severe TBI patients, but the most recent data suggest that blood transfusion itself, rather than the actual Hb value, is associated with a worse long-term functional outcome. Several studies investigating the association of anemia with outcome in patients with severe TBI have used Hb of 10 g/dL to define “anemia,” although a few studies used 8 g/dL and one used 9 g/dL. Thus the appropriate trigger for these patients is unclear, and monitoring of cerebral oxygenation may be a more appropriate end point.

If surgery is indicated, special care should be taken with the ventriculostomy drain; both failure of drainage and excessive loss of CSF can occur during transport. Familiarity with advanced tissue oxygenation monitors, such as those measuring brain tissue oxygen delivery (Pbr o ₂ ) and oxygen consumption (the jugular bulb) (Sjv o ₂ ) levels can also be beneficial. There is an association of poor outcomes with low Pbr o ₂ (< 15 mm Hg), but it is unclear if higher Pbr o ₂ levels correlate with improved outcomes. Until there is consensus, it is recommended to maintain Pbr o ₂ above 20 mm Hg by decreasing O ₂ demand of the brain (decrease ICP, administer analgesia, treat hyperthermia) or by increasing O ₂ supply to the brain (increase cardiac output, transfuse RBCs).

Positional therapy is used in almost every patient with severe TBI. Elevation of the head facilitates venous and CSF drainage from the cranium, lowering ICP and improving CPP as long as the patient is euvolemic. Pulmonary ventilation/perfusion (V/Q) matching may also improve in this position, facilitating maintenance of cerebral O ₂ delivery. The patient should be transported to the OR in this position and maintained with the head up during surgery if possible.

Even in patients with severe TBI, analgesics are indicated for pain arising from coexisting injuries. Sedatives are useful for control of elevated ICP but may make serial examination difficult. Propofol is popular because it offers the most rapid return of neurologic function when discontinued, but the clinician must use this drug cautiously. Large doses of propofol sustained over days to weeks have recently been associated with the development of lethal rhabdomyolysis, the propofol infusion syndrome. This syndrome is more common, and should be suspected, in younger patients, those with severe neurologic injuries, and those who are receiving exogenous vasoactive infusions. The use of sedatives to decrease ICP frequently mandates the use of vasoactive drugs to maintain MAP. Invasive hemodynamic monitoring with a central venous or pulmonary artery (PA) catheter, along with frequent assessment of lactate, base deficit, cardiac output, systemic vascular resistance (SVR), and central or mixed venous oxygen saturation (Sv o ₂ ) may be necessary to maintain appropriate intravascular volume in the presence of the confounding parameters of ongoing shock physiology, pharmacologic agents, and mechanical ventilation.

Osmotic diuretic agents are common first-line agents for severe TBI. Mannitol decreases ICP by drawing edema fluid out of brain tissue and into the circulation. Mannitol may also have a secondary benefit as a scavenger of free radicals and other harmful inflammatory compounds. Hypertonic saline has a similar osmotic effect on the brain, aids in the repletion of circulating volume, and may also act as a beneficial immunologic agent. Use of mannitol or HTS will lead to increased diuresis, necessitating greater attention to adequate volume replacement so that euvolemia can be maintained. Use of osmotic agents to reduce elevated ICP is usually titrated to a serum osmolarity of about 310 to 320 mOsm/L.

Invasive physiologic monitoring, positional therapy, sedation, and osmotic diuresis apply to most patients with severe TBI, but the next tier of therapy is reserved for the subset with intractable ICP elevation. A small percentage may respond to barbiturate coma, which not only lowers cerebral metabolic rate but also decreases excitatory neurotransmitters. Management of barbiturate coma necessitates rigorous attention to intravascular volume, usually requiring a PA catheter and the use of vasoactive and inotropic agents to maintain CPP.

Decompressive craniectomy is gaining increasing acceptance in the management of intractable ICP elevation. Relieving pressure by removal of a piece of cranium and closure with a dural patch may improve outcomes in patients who might not otherwise survive. Randomized studies for decompression therapy after TBI have required years of subject recruitment, and although some show decreased ICP, mechanical ventilation, and ICU stay, long-term outcome (at 6 months) is not improved. Craniectomy will likely remain the procedure of choice for mass lesions, but for diffuse injury, it is still not known which patients will and will not benefit from decompression. Decompressive laparotomy may also be indicated in patients with severe TBI if coexisting injuries or vigorous volume infusion have increased intra-abdominal compartment pressure to greater than 20 mm Hg, a level likely to have adverse effects on intrathoracic, inspiratory, and intracranial pressures.

Although vigorous avoidance of fever is an undisputed recommendation, deliberate hypothermia to reduce the cerebral metabolic rate remains controversial and is not currently recommended. Corticosteroid therapy for severe TBI has not proved beneficial and is now contraindicated because of its high potential for deleterious side effects.

More recently, the clinical picture of “sympathetic storm” (also known as “brain storming”) has been described. Typically seen in younger patients with more severe TBI, but possible at all ages, “storming” is caused by massive catecholamine release. This was initially recognized in patients with nontraumatic subarachnoid hemorrhage (SAH), but storming has since been appreciated in patients with TBI. Patients with severe TBI manifest a hyperadrenergic state with adrenal release of catecholamines and clinically as tachycardia, hypertension, tachypnea, mydriasis, and diaphoresis. They may have a greater than sevenfold increase in norepinephrine, epinephrine, and metabolites. Most pronounced after the first week of injury, treatment of storming in patients with TBI consists of organ system support and may require extreme measures, including extracorporeal circulation. TBI should not be considered to be a contraindication to extracorporeal modalities, as long as exquisite attention is paid to the risks of bleeding and local anticoagulation of the circuit.

A less aggressive treatment available in all hospitals is the use of beta-adrenergic blockade to decrease the sympathetic outflow and to mitigate symptoms. A retrospective review studied trauma patients with an abbreviated injury score (AIS) of 3 or greater who received β-blockade over a 14-month period. “Beta-blocker exposure” was defined as having received β-blockers for 2 or more days. Of the 420 study patients, the 174 who received β-blocker therapy had a slightly higher injury severity, with predicted survival of 59.1% and actual survival of 94.9%; patients who did not receive β-blockers had predicted survival of 70.3% and actual survival of 80.2%. Therefore the patients who received β-blockers, despite their severe injuries, had 5.1% mortality, versus 10.8% mortality in those who did not receive β-blockers. Randomized prospective trials are needed to further elucidate this treatment.

Multiple-organ failure associated with neurologic injury is increasingly recognized as a sequela of the initial insult. Neuroinflammation may be an important mediator of secondary injury; patients with TBI have elevated CSF cytokine levels, with systemic delivery of these cytokines. These inflammatory mediators probably play a large part in the development of nonneurologic organ dysfunction. Of 209 consecutive multiple-trauma patients with severe TBI who required more than 48 hours of intensive care management, 89% developed dysfunction of at least one nonneurologic organ system. Respiratory dysfunction was most common (81%), followed by cardiovascular (52%). Although seen in smaller percentages of patients, hematologic (36%), hepatic (8%), and renal (7%) dysfunction were also present. More importantly, hospital mortality in this study was associated with organ system failures: 26% for patients without nonneurologic organ system failure, 40% for those with one organ system failure, 47% for two failures, and 100% in the small proportion of patients with three or more nonneurologic organ failures.

Spinal cord injury (SCI) is also associated with multiorgan failure. In a retrospective review over 15 months, of 1028 patients admitted with SCI, 40 were identified with isolated injury and ICU stay longer than 24 hours (AIS > 3, with other organs excluded). “Organ failure,” defined as failure in at least one organ system, occurred in 75% of patients by multiple-organ dysfunction score (MODS) criteria and 85% of patients by sequential organ failure assessment (SOFA) scoring. There was an inverse correlation between the American Spinal Injury Association (ASIA) score, which defines the motor and sensory level of injury, and MODS/SOFA scores. Patients with more severe (higher level or complete) SCI may therefore benefit from the specialized care of traumatic SCI units and rehabilitation centers.