Critical care of the multiple-injury patient

Seminars in ANESTHESIA, PERIOPERATIX E MEDICINE AND PAIN Vol 18, No 3

September 1999

Critical Care of the Multiple-Injury Patient Michael J. Sullivan N ENORMOUS body of literature is devoted to critical care. This article is intended for the practicing anesthesiologist not routinely involved in the care of critically ill patients. The specific patient type is the acutely injured patient in hypovolemic shock. The focus is on the physiological changes associated with hypovolemia, preoperative operating room preparation, assessment of the patient with multiple injuries, intraoperative resuscitation, and avoidance of postoperative complications.

RESPONSE TO HYPOVOLEMIA Neuroendocrine Response

The aortic arch and carotid bodies sense the decrease in pressure with acute blood loss. The compensatory response is to maintain perfusion to the central nervous system and the cardiovascular system. To achieve this purpose, peripheral vasoconstriction and fluid retention occur. The mechanisms involved are activation of the sympathetic nervous system, hormonal responses that augment the sympathetic response, and local microcirculatory responses that are organ specific and regulate regional blood flow. The magnitude of the response is foremost related primarily to the rapidity and amount of volume lost and secondarily to the underlying physiological reserves and to the presence or absence of chemicals, both therapeutic and recreational. Sympathetic nervous system vasoconstriction is an immediate response. Examination of the distribution of the capacitance vessels shows that the

majority of blood (65%) is on the venous side. The body draws on this volume in its time of need, and peripheral capacitance is reduced. Peripheral perfusion is sacrificed to central perfusion. Sympathetic activation of the adrenal medulla occurs with the release of epinephrine and norepinephrine, causing systemic effects. They increase blood pressure, heart rate, cardiac contractility, and minute ventilation. They modulate the handling of energy substrates such as glucose, amino acids, and fatty acids. They augment sympathetic arterial and venous vasoconstriction. Arterial vasoconstriction is not distributed evenly. Flow to the heart and brain are maintained until all compensation fails. Local endothelial vasoregulation of the heart and brain maintains regional perfusion despite intense sympathetic and catecholamine responses. This compensation does not require factors extrinsic to the microvascular bed. Blood flow to other, nonvital organs markedly decreases. Hypoperfusion of the brain induces the vasomotor center of the medulla to increase an already elevated sym-

From the Los Angeles County~University of Southern California Medical Center, Los Angeles, CA. Address reprint requests to Michael J. Sullivan, MD, Los Angeles County/University of Southern California Medical Center, 1200 N State St, Suite 14-902, Los Angeles, CA 90033. Copyright 9 1999 by W.B. Saunders Company 0277-0326/99/1803-0001 $10.00/0

Seminars in Anesthesia, Perioperative Medicine and Pain, Vol 18, No 3 (September), 1999: pp 17"7-191

177

178

MICHAEL J. SULLIVAN

pathetic output in a final effort to perfuse the brain. Vasoconstriction of the venous side can result in increased intraluminal capillary pressure, swelling the capillaries. This effect is more pronounced when resuscitation increases the arterial pressure and venovasoconstriction is still present. The hydrostatic and oncotic forces (Starling forces) across the capillary bed normally favor filtration) -4 Even a small increase in the hydostatic pressure can markedly increase filtration. Net filtration = Kf[(Pcap - Pif)

-

-

(lip - Ilif)],

where Kfrefers to the net permeability of the capillary wall, Pcap(17 mm Hg) and Pie (4 mm Hg) are the capillary and interstitial hydrostatic pressures, and lip (26 mm Hg) and I~i"f (5.3 n'l_n3Hg) are the plasma and fluid oncotic pressures. When average values obtained from peripheral capillaries such as skeletal muscle are used, the equation becomes [ ( 1 7 - { - 4 } ) - (26 - 5.3)] = +0.3. Other tissue beds such as the liver and kidneys have different hydrostatic and oncotic pressure values. The net effect in these capillaries is a filtration value even greater than that in the periphery.5 This causes loss of intravascular volume to the interstitium or intracellular compartments with tissue edema formation. The endocrine response is the release of multiple hormones from the hypothalamic-pituitary-adrenal axis whose end result is fluid retention. Corticotrophin-releasing factor from the hypothalamus stimulates the release of adrenocorticotropic hormone (ACTH) from the anterior lobe of the pituitary, whose target organ is the adrenal cortex. Cortisol secretion is central to the stress response. The degree of hypercortisolism parallels the degree of injury and has been used as a marker for the degree of the stress response. Its effects include sodium retention, insulin resistance, gluconeogenesis, and lipolysis. Along with its own effects on protein catabolism, it enhances the catabolic effects of tumor necrosis factor and intefleukin 6. The posterior lobe of the pituitary gland releases antidiuretic hormone. Its dual actions are promoting water reabsorption at the renal collecting ducts and acting as a pressor agent. It has potent vasoconstrictor effects. Other hormones released include glucagon and growth hormone. These hormones oppose the effects of insulin by promoting gluconeogenesis, lipolysis, and glycogenolysis. The tissues themselves show increased resistance to the action of insulin. 6

This hyperglycemic statehelps to pull fluid from the interstitium into the intravascular space and maintain circulating volume. The juxtaglomerular cells of the kidney release renin, which triggers the renin angiotensin aldosterone system. The overall effect is vasoconstriction and a decrease in salt and water excretion by the kidneys. Local mediators are produced when endothelial injury is caused by ischemia or disruption. They have both local and systemic effects. Exposure of subendothelial collagen and basement membrane activates circulating Hageman factor (factor XII). This activates the intrinsic pathway of the clotting system. Factor XII initiates an inflammatory response by activating the complement cascade. Depending on the magnitude of injury and the vigorousness of the body' s response, the entire systemic inflammatory response syndrome (SIRS) can occur. Macrophage function is affected, oxygen-derived free radicals are produced, and the multiple products of arachidonic acid are synthesized (prostaglandins, thromboxanes, leukotrienes). Cytokines are elaborated, and neutrophils and endothelial cells interact together. Activation of inflammation and coagulation at the sites of injury is essential for fighting or prevention of infection and for eventual healing and repair. However, an overvigorous inflammatory response can compromise organ function at sites distant from the injury. The neuroendocrine response is a systemic response that involves modulation of the cardiac, pulmonary, and arteriovenous tissues. Sympathetic activation initiates the release of catech01amines and multiple hormones. Local tissue injury caused by mechanical disruption, ischemia, or reperfusion initiates an inflammatory response that has a beneficial purpose but can be deleterious.

OPERATING ROOM PREPARATION Preoperative preparation of the operating room is essential. If time allows an evaluation of the patient before arrival in the operating room, specific setup of equipment and medications is possible. If the operating room is to be prepared for "what ever comes through the door," the following setup is necessary. Preparation begins with a machine check, ensuring the ability to administer positive pressure ventilation by mask or endotracheal tube, performing a highpressure check for leaks, and presetting the ventilator

THE MULTIPLE-INJURYPATIENT for a standard 70-kg patient. A tidal volume of 700 cm 3 at a rate 10 breaths/min, with an I:E ratio of 1:2, will safely ventilate a small female to a large male patient. In the hectic minutes after intubation is performed and before all of the anesthetic lines are in, these ventilator settings will free up attention for other critical procedures until it can be returned for finer adjustments of ventilation. Rapid-sequence induction is the method of choice unless specific problems require a different method. It is important to know where the difficult airway cart is located. A proper-sized styleted endotracheal tube with a J bend in the tube is required. Two different-sized blades of the same style and another blade of a different style, for example a Mac 3, a Mac 4, and a Miller 4, should be available and ready. Two different-sized oropharyngeal airways and a second ventilatory mask should be at arms reach. Rapid availability of a double-lumen tube is nice if there is the possibility that one-lung ventilation will be needed during surgery. It is easier to place this tube at the beginning. A working suction device capable of suctioning large particulate matter whose tubing easily reaches the patients mouth should be available. All acutely injured patients are hypovolemic. Equipment and supplies to rapidly place large bore peripheral intravenous lines, 14 gauge or 16 gauge angiocatheters, and central lines are mandatory. Rapid fluid warmers primed and available to attach to these catheters are also needed. After successful venous cannulations are finished, attention is turned to arterial catheterization. Arterial catheters, pressure tubing, and pressure transducers should be ready. A second pressure transducer for central venous pressure monitoring is recommended. If technical ability allows, time permits, and it is clinically indicated, a pulmonary artery catheter is valuable. A Swan-Ganz catheter is not a therapeutic measure, and the optimal time to use a pulmonary artery catheter is before organ failure begins. Pressure transducers need to be ready for pressure readings from the pulmonary artery catheter. Adjuncts are warming the operating room, turning on heating mattresses, and having warming equipment handy such as Baer Hugger warming blankets. The blood bank should be alerted if they are not automatically notified that blood products may be necessary, and extra intravenous fluid bags should be available in the work space. Finally, an assistant to the anesthesiologist is invaluable.

179 PREOPERATIVE ASSESSMENT OF CRITICALLY INJURED PATIENTS Anesthesiologists and surgeons have classification systems to convey an overall sense of the severity of a patient's underlying physiological status (anesthesia) and the rapidity in which anatomical repair is warranted (surgery). American Society of Anesthesiology (ASA) criteria classifies patients in categories ASA I through ASA V, and the suffix E denotes emergency procedures. Class I indicates a healthy individual, and class V indicates a moribund patient who is not expected to survive the next 24 hours with or without surgery. The American College of Surgeons (ACS) classification incorporates timing of the surgical procedure. Class I indicates that surgery is required immediately if the patient is to survive, class II indicates that surgery is urgent and clearly indicated, and class III indicates a need for preoperative optimization before surgery to minimize postoperative morbidity. These global paradigms basically stratify patients into those with low, intermediate, or high risk of morbidity or mortality. Individual patient assessment of airway, neurological, cardiovascular, pulmonary, renal, and circulatory status are of paramount importance, A thorough physiological assessment often is not possible preoperatively, and intraoperative assessment and management occur simultaneously.

Airway Airway management in the trauma patient is always the first priority. Supplemental oxygen should be administered while the patient is being assessed. If a second person experienced in airway management is available, he or she should be invited, ordered, begged, pleaded with, or threatened into assisting. The ASA difficult airway algorhythm is a valuable source of information. Surgical establishment of the airway is an option that should be considered early. It is best not to walt until every method of tracheal intubation is tried and unsuccessful before surgical intervention is considered. Intubation while the patient is awake is another option if loss of the airway or hemodynamic instability is a concern. Facial injures, cervical spine injuries, and basal skull injuries make securing the airway a high-risk procedure. Fortunately, isolated facial injuries are rarely associated with cervical spine injuries. The vast majority of patients with combined facial and neck injuries have multiple system trauma. In a study of 2,555 patients

180

MICHAEL J. SULLIVAN

with facial fractures severe enough for hospital admission, only 1.3% had associated neck injuries]

Neurological Head injuries. Head injury management is covered in another article is this issue. Nearly 50% of all trauma deaths have associated head injury. The Glascow Coma Scale (GCS) is useful for a quick examination (Table 1). The GCS measures the patient's ability to blink, speak, or move extremities: spontaneous blinking, 4 points; full speaking abilities, 5 points; voluntary movement of extremities, 6 points. GCS score of -<8 is associated with a mortality of approximately 30%. If the GCS is ->9, the mortality is approximately 1%. To assess level of consciousness rapidly, the AVPU (alert, verbal, painful unresponsive) scale is used in response to stimuli. Interventions to prevent secondary injury include tracheal intubation, hyperventilation, and judicious use of isotonic fluids. The use of sedation is always questionable in head trauma because it may affect neurological results of the examination. Administration of sedation and analgesia in an obtunded or combative patient, who often has other significant injuries, can decrease oxygen requirements and may save a few more cells. Agents today are shorter acting and can be reversed if needed.

Table 1. The Glascow Coma Scale Measure of Level of Consciousness

Measure

Response

Eye opening

Opens Spontaneously To verbal command To pain No response Orientated and converses Disorientated and converses Inappropriate words Incomprehensible sounds No response Obeys verbal command To painful stimulus Localizes pain Flexion, withdrawal Abnormal flexion (decorticate rigidity) Extension (decerebrate rigidity) No response

Verbal

Motor

Score 4 3 2 1 5 4 3 2 1 6 5 4 3 2 1

Spinal cord injuries. All multiple-injury patients are assumed to have spinal injury until proven otherwise. Spinal cord injuries can occur anywhere along the spinal canal. Certain injuries have a higher chance of spinal disruption in specific locations than others. At low risk for spinal injuries are patients who are awake, alert, and have no spinal pain. Moderate risk is found in injuries associated with motor vehicle accidents, contact sport injuries, direct head injuries, and falls. Highrisk injuries involve front-end motor vehicle accidents and head-first falls. The major focus is on the cervical spine. Approximately 39% of cervical fractures have associated neurological injuries. Missing or mishandling a cervical spinal injury is catastrophic. This is an important consideration for the anesthesiologists who is attempting to secure the airway. The rigid collar prevents approximately 70% of flexion/extension motion and 50% of rotational motion. Clinical clearance is safe if the following criteria are met: (1) no neck pain or tenderness; (2) no transient or persistent numbness or paresthesia; (3) no loss of consciousness; (4) no impaired level of consciousness from alcohol or drugs with evidence of head or neck injury; (5) absence of physical findings of cervical cord injury peripherally, eg, anterior spinal cord syndrome; and (6) no other painful distracting injuries, especially multiple injuries resulting from high-energy accidents or falls. Radiographic clearance involves lateral, anterior posterior, and open-mouth odontoid views of the cervical spine. These three views can rule out approximately 99% of cervical bony injuries. Cervical vertebral bodies 1 through 7 must be visualized. A lateral view alone has a sensitivity of approximately 85% for significant bony injuries. Rarely do these x-rays arrive with the patient, and even if they do, anesthesiologists are uncomfortable being the ones to "clear the C-spine." If a spinal injury is suspected, the patient is moved as one complete unit, avoiding neck flexion, extension, or rotation. Airway manipulation proceeds with inline stabilization of the cervical spine. Pharmacological care of the injured spine begins with the administration of steroids. Methylprednisolone is given within 8 hours of injury in a bolus dose of 30 mg/kg, followed by 5.4 mg/kg/h for the next 23 hours. Improvement has been noted at 6 months in motor function and pinprick and touch sensation. The fourth cervical vertebrae is important. Spontaneous ventilation is possible in cord injuries

THE MULTIPLE-INJURYPATIENT below this level because diaphragm innervation is intact. Low spinal cord injuries cause sympathectomy. Patients are vasodilated below the level of the injury, causing low blood pressure and a low systemic vascular resistance, a-Agonists can be used when heart rate and cardiac contractility are not problematic. Chest Trauma

The organs at risk for injury are the ribs, lung, heart, esophagus, and major vessels. Twenty-five percent of trauma deaths are caused by chest injuries. Rib fractures are the most common thoracic injury. Multiple rib fractures or fracture of T1 or T2 involves greater energy transfer to the patient and the possibility of associated injuries. The most common intervention required with a fractured rib is placement of a chest tube to release blood, air, or both from the chest cavity. If more than 1,500 mL of blood initially or more than 200 mL/h continuously drains from the tube, surgical intervention in the chest is warranted. Flail chest involves three adjacent ribs broken in two places. A section of the thoracic cage can be removed without further cutting of ribs. Management currently involves pain control and monitoring of pulmonary function. Postoperative pain management in chest trauma improves pulmonary function. Pulmonary evaluation involves searching for risk factors of lung compromise such as tobacco use, dyspnea on exertion, chronic obstructive pulmonary disease (COPD), environmental exposure to pollutants, asthma, morbid obesity, neuromuscular disease, and systemic diseases with pulmonary manifestations (acquired immunodeficiency syndrome, tuberculosis, systemic lupus erythematosus, and sarciodosis). Physical examination searches for changes associated with COPD (increased AP diameter, expiratory wheezing, and hyperresonance). If the patient arrives from the emergency department intubated, the chest x-ray may reveal valuable information along with assessment of ventilatory parameters after the anesthetic machine is connected. The Fio2 needed to maintain arterial saturation, the peak inspiratory pressure, the capnographic waveform pattern, and positive end-expiratory pressure give a global measure of the ability of the lungs to oxygenate and ventilate the patient. Much like preoperative pulmonary function test results, there is no correlation be-

181 tween these values and the incidence of postoperative pulmonary complications. Pulmonary contusion is managed by the clinical status of the patient. This means possible intubation. Risk factors that have been shown to increase the likelihood of intubation are initial respiratory rate greater than 25 breaths/min, pulse rate grater than 100 beats/min, systolic blood pressure less than 100 mm Hg, and the presence of other injuries. The lung suffers damage because of direct trauma or remote organ ischemia-reperfusion that affects lung function. If the damage is severe enough, is decreased ability to oxygenate blood an increased Fio 2 are seen. This is acute respiratory distress syndrome (ARDS), which occurs most frequently in postoperative or posttraumatic patients who may have had hypotension, unstable hemodynamics, or shock during or shortly after surgery or trauma and in whom fluid therapy has restored blood pressure and urine output. Treatment of ARDS is beyond the scope of this article. It begins in the intraoperative period but is usually not a significant problem until later in the hospital course. Aspiration of stomach contents into the pulmonary system can occur at any time. Critical points are induction of and emergence from anesthesia. Methods to lessen the risk in trauma patients include rapid-sequence induction with criciod pressure and placement of a cuffed tube in the trachea. Pulmonary aspiration is responsible for up to 20% of all anesthetic-related deaths. Trauma patients are at particularly high risk. They are considered to have full stomachs because gastric emptying ceases at the time of injury. Limitations on the ability to have the proper sniffing position alignment (C-spine collar still in place) for intubation adds to the risk. The type and volume of the aspirate affect the virulence of the injuries to the lungs. Morbidity increases with a pH below 2.0 to 2.5 and a volume more than 0.4 to 1.0 mL/kg. Nonacid liquids such as blood and saliva can cause laryngospasm, brochospasm, and loss of surfactant but little histological abnormality in the lungs. Acid liquid aspiration (pH < 2.5) causes atelectasis, alveolarcapillary breakdown with hemorrhage, and interstitial edema. Pulmonary hypertension and hypoxia can occur early. Particulate aspiration can cause an inflammatory response, and granulomas may form around the particulate nidus. Par-

182 ticulate antacids should be avoided. Trauma to the oral cavity can dislodge teeth, which can be aspirated. Clinical signs of aspiration include vomitus in the oropharynx, wheezing, cyanosis, coughing, hypoxemia, pulmonary edema, and hypotension. Chest x-ray findings can range from no changes initially to diffuse bilateral infiltrates. If aspiration does occur and time permits a rapid-sequence intubation, catheter suctioning of the endotracheal tube before administration of positive-pressttre ventilation will help prevent further spread of the contents. If the apneic oxygenation time has elapsed, adequate oxygenation and ventilation should be assured, then attempts to clear large aspirated matter are made. Bronchoscopy usually is not indicated unless significant solid material has been aspirated. Prophylatic antibiotics are not recommended, nor is steroid administration. Antibiotics are used if there is a secondary bacterial pneumonia or suspected aspiration of fecal material. Airway management takes precedent over aspiration. Critically injured patients can have underlying cardiac disease. The reason for a motor vehicle accident may be acute myocardial infarction or a cardiac ischemic event. In these cases, the cardiac risk factors defined in the Guidelines for Perioperative Cardiovascular Evaluation for Noncardiac Surgery should be considered, although surgery will still proceed. 8 Classic risk factors include congestive heart failure, recent myocardial infarction, major arrhythmia, severe angina, hemodynamically significant valvular disease, and severe hypertension. A median sternotomy scar from a previous CABG should be looked for. Factors associated with increased mortality for noncardiac surgery after CABG are prior myocardial infarction, high left ventricular score, preoperative use of nitrates, congestive heart failure, male sex, diabetes, age, dyspnea on exertion, and left ventricular hypertrophy seen on the electrocardiogram. Ischemia can be identified on the intraoperative cardiac monitor. Nitroglycerine can be used safely in a critically injured patient if invasive monitoring of blood pressure is carried out. Abdominal Trauma

Abdominal trauma is the most frequent cause of treatable early life-threatening hemorrhaging in injured patients. Exsanguination from organ injury varies depending on the organ injured. Death by

MICHAEL J. SULLIVAN exsaguination in abdominal aortic injury is approximately 50%, whereas it is approximately 6% in hepatic injury. The most common injury requiring abdominal intervention in blunt trauma is that of the spleen. Serial abdominal examination is the best clinical test for abdominal pathology in a cooperative patient. In unreliable patients, diagnostic peritoneal lavage and computed tomography (CT) scanning are two common modalities for abdominal examination. Each has its limitations, but usually only one is used. CT scanning requires a patient stable enough to travel to the radiology department. In acutely injured patients, the anesthesiologist has little preoperative information about specific intraabdominal organs at the time of surgery. The decision is usually whether the abdomen warrants surgical exploration or not. If warranted, the surgeon visually examines each organ. A brief description of the surgical management of organ injuries and the organs' responses to hypovolemia follows. Liver. Treatment of liver injuries ranges from nonsurgical management to perihepatic packing and planned reexploration when greater hemodynamic stability is possible and coagulopathy is treated. The liver is susceptible to hypoperfusion and ischemia. A reduction in hepatic oxygen supply with increased oxygen demands has been postulated to contribute to hepatocellular failure. The magnitude of the oxygen debt to the liver and subsequent dysfunction is open to debate. The liver receives approximately 1.5 L of blood flow per minute. This is from two sources. Portal venous blood flow provides 60% to 70% of total blood flow and 50% to 60% of the oxygen used by the liver. The hepatic artery supplies 30% to 40% of the hepatic blood flow and approximately 40% to 50% of the oxygen. The hepatic artery is able to increase flow by decreasing resistance when the oxygen needs of the liver are not being met. This occurs when portal blood flow is decreased or when mixed venous oxygen saturation is decreased, as in the hypovolemic state. Other factors reduce hepatic blood flow. These include mechanical ventilation, positive-pressure ventilation, reduced cardiac output, hypoxia, decreased mixed venous oxygen content, systemic hypotension, and hypocarbia. To maintain liver viability, flow must be restored. In a study by Chun et al, 9 sinusoidal blood flow and hepatocyte viability were both preserved with flow-controlled reperfusion as opposed

THE MULTIPLE-INJURYPATIENT to pressure-controlled reperfusion, emphasizing the critical role of adequate microvascular flow in the maintenance of hepatic viability. Restoration of normal vital signs may not render adequate tissue per fusion. Pancreas. Pancreatic injuries are approximately 12% of the injuries to victims of abdominal trauma. Two thirds of these injuries are penetrating in nature. Mortality from pancreatic trauma ranges from 12% to 30% and often occurs early in the clinical course. Late death results from uncontrolled pancreatic sepsis caused by inappropriate initial management or delayed management of complications. These complications include intraabdominal abscesses, fistulas, pseudocysts, and pancreatic insufficiency. Surgical management usually involves debridement or excision of pancreatic tissue. Because of its retroperitoneal location and its proximity to major vascular structures, associated injuries play a significant role in morbidity and mortality. Pancreatic injuries usually do not cause difficulty for the anesthesiologist in terms of resuscitation or blood loss at the time of the initial insult. Because the pancreas is a retroperitoneal structure, diagnostic peritoneal lavage may miss evidence of organ damage. Amylase and lipase determination are of limited usefulness. This organ is subjected to a severe reduction in blood flow during hypotension, possibly to only 15% of baseline blood flow. '~ This deficit often persists despite adequate volume replacement. The magnitude of the effect of hypoxia and hypoperfusion upon pancreatic function is unclear. Stress and shock cause the release of a number of hormones that affect the pancreas. Catecholamines stimulate glucagon release and decrease insulin secretion, thus augmenting hepatic gluconeogenesis and the initial hyperglycemia associated with hemorrhagic shock. Spleen. The spleen is the most frequently injured organ in blunt abdominal trauma and is commonly associated with other intraabdominal injuries. Splenectomy was the normal treatment until the risk of postsplenectomy sepsis resulting in morbidity and mortality was appreciated. Currently splenorrhaphy, nonsurgical observation, and splenectomy are the standard treatments. Clinical findings in splenic injury include those seen with blood loss: tachycardia, hypotension, and syncope. Left shoulder strap pain, known as Kehr's sign, is a classic finding in splenic rupture, left upper quad-

183 rant abdominal pain is usually elicited, and fractured ribs of the lower left thoracic rib cage indicate possible splenic damage. Little is known about splenic hemodynamics during and after hemorrhage shock. The duration of hypotension and hypoperfusion and its effects on organ function and remote organ systems are not well characterized. Intuitively, reestablishing flow as soon as possible seems prudent. Intestines. Small bowel injuries are usually the result of penetrating injury and generally are associated with a good prognosis. Injury to the small bowel results in spillage of contents and early peritoneal irritation with significant symptoms. Blunt trauma requires a higher index of suspicion because evaluation of these injuries can be difficult. CT scan and diagnostic peritoneal lavage resuits can both be negative. The bowel can lose its viability over time. The majority of bowel injuries can be repaired primarily after local debridement. More extensive injuries or loss of the blood supply to various segments are repaired by resection and anastomosis. The role of the gut in shock is that of a depot of bacteria that can translocate into the systemic circulation. This failure of the gut barrier function has been hypothesized as a contributor to multiorgan dysfunction and failure, trauma, and hemorrhagic shock. Experimental studies in animals have found conflicting results. The clinical significance of this translocation is unclear. During intestinal ischemia, metabolites are produced that can cause lung damage when perfusion is restored. This lung damage has been shown even in the absence of bacterial translocation. The response of the gut to hemorrhaging is caused by vasoconstriction, which diverts blood flow away from splanchnic tissue. Angiotensin II and vasopressin are the major intestinal vasoconstrictors, with help from increased sympathetic tone and thromboxane A 2 and thyrotrophic-releasing hormone levels. Not all splanchnic vessels constrict. Large arterioles constrict, and smaller, premucosal arterioles dilate. 13 Intestinal hypoperfusion persists even with restoration of normal blood pressure. 14-17 Damage to the endothelial cells, endothelial cell swelling, thombosis, and microvascular injury are possible explanations for continued hypoperfusion with normal blood pressure. Genital-urinary system. Blunt trauma is responsible for one of the most common renal inju-

184 ries, renal contusion, which usually resolves without intervention and without long-term morbidity. Urir,e dipstick results usually are positive for hematuria. The degree of hematuria does not correlate with the extent of injury. Contrast studies are needed for persistent or gross hematuria. Contrast CT provides excellent information if time allows. Surgical repair involves partial or total nephrectomy of a damaged kidney. Ureters and bladder are repaired if possible and resected if necessary. In the resuscitation of a hypotensive patient, the ability to monitor urine output is paramount. Knowledge of the functionality of the system is important. Blood at the ureteral meatus, scrotal hematoma, or a high-riding prostate necessitates retrograde urethrography before Foley catheter insertion to ensure urethral patency. A suprapubic catheter can be placed if there is urethral disruption. In emergent situations, bladder and kidney evaluation can be performed by rapid cystogram and intravenous pyelogram (IVP). Oblique, lateral, and postvoid views increase the chance of finding subtle injuries. In the hypotensive patient, nonvisualization of IVPs is common and should await patient stabilization. Intraoperative studies are technically inferior but identify extravasation or absence of function. Dyes that are excreted in the urine, such as methylene blue and indigo carmine, can also be used to identify locations of extravasation. Hemorrhagic shock results in decreased renal blood flow. Avoidance or prevention of acute renal failure is essential for overall recovery from hypovolemic shock. The kidneys increase efferent arteriolar tone more than afferent arteriolar tone. This helps maintain glomerular filtration rate. Regional differences occur in blood flow, superficial renal cortical blood flow is decreased more than deep cortical blood flow and medullary blood flow. If these compensatory mechanisms fail or are insufficient to maintain renal perfusion, patchy tubular epithelial injury occurs in the already impaired microcirculatory environment. The magnitude of renal injury correlates with the occurrence of complications, renal failure, and death. The most common causes of renal failure are hypoperfusion and nephrotoxic drugs. Rhabdomyolysis from crush injuries, extensive tissue injuries, and ischemia also occur in patients with multiple injuries. Decreased renal perfusion can be caused by increased abdominal pressure. Renal

MICHAEL J. SULLIVAN paranchymal and venous outflow compression induce renal dysfunction. Tissue edema, bowel edema, continued blood loss, ascites, laparotomy packs, and liver packing are all contributing factors to pressure-induced renal ischemia. Maintenance of perfusion and volume resuscitation are essential, as in all organs. Mannitol has renal-protective effects in preservation of kidney function. Trauma to the Pelvis and Femur

Pelvic fractures comprise 3% of all skeletal fractures. These fractures and associated injuries are a frequent cause of death from blunt trauma. Most pelvic fractures are secondary to automobile or pedestrian accidents. Bleeding can be massive from pelvic fractures, and the mortality rate associated with open pelvic fractures is as high as 50%. Most of the bleeding is venous in origin. Hemorrhage is the major cause of death in pelvic injuries. Retroperitoneal bleeding always occurs. Up to 6 L of blood can be shed into this space. Because of the difficulty of controlling this source of bleeding, avoidance of opening a retroperitoneal hematoma is advocated. Bleeding occurs from both large and small vessels, especially the superior gluteal and internal pudental branches of the internal iliac artery. Failure to restore hemodynamic stability within a short time is indirect evidence of arterial bleeding, which can be controlled with selective angiography and embolization. Use of arteriography for the treatment of pelvic fractures is controversial. Studies show that 10% to 15% of pelvic fractures involve arterial bleeding. Many surgeons believe that all hemodynamically unstable patients with pelvic fractures should routinely undergo angiography immediately after they leave the emergency department. This can be a life-saving intervention, although it does not control venous bleeding. Regardless of the preceding evaluations and therapy, anesthesiologists usually encounter hemodynamically unstable patients with multiple injuries who are in need of resuscitation. Radiographic evaluation. Stabilizating the patient takes precedent over obtaining x-rays. Unnecessary movement may produce further injury and blood loss. In an unconscious patient with multiple injuries, an AP view of the pelvis is a necessary baseline. Additional studies include lateral, inlet, outlet, and internal and external oblique pelvic views. CT scans and special studies may be needed to evaluate acetabular and sacral fractures fully.

THE MULTIPLE-INJURYPATIENT

185

Angiography and venography may be necessary to determine the source of bleeding. Each patient's condition must dictate what is done and when. Treatment. Pneumatic antishock trousers are used to help stabilize the pelvis and decrease bleeding. Use of this garment is controversial. It may be helpful to control bleeding. Disadvantages include decreased access to the abdomen and extremities and compartment syndrome with prolonged use. When faced with a patient with this garment, one should slowly deflate each compartment while monitoring blood pressure. If hypotension occurs, a decision between more fluid resuscitation and reinflation of the garment must be made. Anatomic stabilization or fracture reduction and fluid restoration are necessary. Placement of an external fixator stabilizes the bone fragments and can dramatically decrease blood loss. Early use of this device has been shown to reduce the incidence of ARDS. Unfortunately, placement of the external fixator is often delayed pending diagnosis and therapeutic interventions in higher-priority organ systems such as the cranial, thoracic, and abdominal cavities. If an exploratory celiotomy is required, application of the external fixator is delayed in the belief that it poses an obstacle to adequate exploration. Without the external fixator, opening the abdomen removes one of the last soft tissue restraints and further increases the external rotation of the unstable hemipelvis. At this point, further hemorrhaging occurs. Mechanical compression and tamponade are often ineffective to control bleeding, and the patient dies. As anesthesiologists we desperately attempt to replace the fluid in a tank in which the hole is larger than our resuscitation hose. Definitive treatment is usually open reduction and internal fixation of the pelvis. This may occur primarily or after stabilization of the patient and a return trip to the operating room once other lifethreatening injuries are controlled.

first 24 hours. Evaluation of the blood supply to the traumatized extremity must be assessed. Peripheral pulses are examined. If they are absent, capillary flow is assessed. The refill time is normally 3 seconds or less. Good capillary flow can ensure survival of the extremity in the absence of pulses. Doppler ultrasonography and arteriography are methods of evaluation of peripheral flow when in doubt. Compartment syndrome should be watched for in a traumatized extremity that is receiving benign neglect as other, higher-priority injuries are being repaired in the operating room. Early diagnosis and treatment is curative, and delays result in permanent and severe disability. Prolonged compartment syndrome complications include muscle or nerve injury, rhabdomyolysis, and loss of the extremity. The pain associated with a developing compartment syndrome is not vocalized by the patient who is under general anesthesia. In an awake patient, palpation of the affected extremity, passive stretching of the muscles, or active contraction of the muscles will elicit pain. Normal compartment pressure ranges from 0 to 8 mm Hg. Capillary blood flow is compromised at pressures greater than approximately 20 mm Hg. Pressures of 30 to 40 mm Hg can cause ischemic necrosis of muscle and nerves. Muscle is more sensitive than nerves to the effects of increased compartment pressure. Blood flow in the arteries, arterioles, and collaterals still occurs at these pressures, although perfusion of the tissues is severely compromised. When distal pulses are noted to be reduced, muscle necrosis has occurred. Any muscle mass invested with fascia is at risk. The same vigilance in monitoring limb position and tourniquet time will ensure adequate assessment for developing compartment syndrome in the anesthetized patient.

Extremity Fractures Extremity fractures usually are not life threatening and are a second-tier concern in the patient with multiple injuries. Early stabilization of these fractures can significantly decrease morbidity. There is less incidence of ARDS, fat embolism, and pneumonia, and fewer days are spent in the intensive care unit after stabilization within the

but sensitive for shock. They can be influenced by pain, anxiety, temperature, therapeutic medications such as beta blockers, and illicit drugs. Some patients have physiological reserves that allow them to maintain vital signs in the normal range until terminal cardiovascular collapse. Vital signs still are the most commonly used parameters in assessment of adequacy of resuscitation. In certain pa-

INTRAOPERATIVE MANAGEMENT Clinical and Laboratory Markers of Acute Hypovolemic Shock Vital signs. Altered vital signs are nonspecific

186 tients, normalization of vital signs is all that is needed. A decrease in heart rate, an increase in blood pressure, a decrease in the Fio 2 needed to maintain an adequate Pao 2, an increase in urinary output, and longer time intervals between therapeutic interventions signals a movement toward normal homeostasis. Clinically this can be described as pulse less than 100 beats/min, pulse pressure greater than 30 mm Hg, urine output greater than 0.5 to 1.0 mL/kg, absence of metabolic acidosis, and minimal effects of positive-pressure ventilation. These are the most basic, familiar, and common end points used as markers for shock and resuscitation. Base deficit. Easily calculated from arterial blood gas analysis, the base deficit is the sum of all metabolic acids including lactate caused by hypoperfusion and ischemia. Base deficit can be an accurate predictor of mortality in patients with multiple trauma. The magnitude of the initial base deficit is a reliable early indicator of the severity of the volume deficit. The amount of fluid required for resuscitation is greater in patients with more severe base deficit than in those with less severe base deficit. In one study, ongoing hemorrhaging was suggested by a progression in the base deficit in 65 % of patients with persistent bleeding. 18 Base deficit is easy to use and obtain from arterial blood gas measurement. Its limitations are that the use of bicarbonate during resuscitation and any preexisting medical conditions that result in chronic elevation or reduction in bicarbonate levels falsely skew the results. Serum lactate level. Lactate level is a marker for anaerobic metabolism. The amount of lactate produced is believed to correlate with the total oxygen debt, the magnitude of hypoperfusion, and thus the severity of shock. When lactate levels normalized within 24 hours of injury, patients reported by Abramson and Scalea 19 had a 100% survival rate. High serum lactate levels have been associated with high mortality among critically ill patients. However, as a marker lactate has its limitations. Some patients experience resolution of lactic acidosis, but it is not always accompanied by improvement in systolic blood pressure or survival. Lactate is cleared by the liver; therefore, injury to the liver or hepatic disease can decrease lactate clearance, leading to high levels not associated with ongoing tissue hypoxia. Tissues that are not perfused during true ischemia do not contribute to

MICHAEL J. SULLIVAN lactate levels measured in the laboratory. Lactate offers no information on regional distribution of tissue hypoxia and blood flow. The belief that an increase in arterial lactate concentration indicates organ ischemia is simplistic. Oxygen transport parameters. The critical molecule transported by the cardiovascular system is oxygen. It has been found that total oxygen debt and rate of accumulation of the debt are both critical determinants of survival. Guyton 3 measured oxygen consumption before and during experimental hemorrhagic shock in dogs. Survivors had oxygen debt less than 100 mL/kg, the LDs0 was 120 mL/kg, and nonsurvivors had oxygen, debt of 140 mL/kg. How do we measure how much oxygen is being used by the body, and how do we improve delivery of that oxygen to the tissues? Simplistically, measuring the oxygen content difference between arterial blood leaving the left ventricle and venous blood leaving the right ventricle tells us the amount of oxygen consumed by the tissues being perfused. Arterial-venous oxygen content difference and cardiac output can be measured using a Swan-Ganz catheter and an arterial line. Shoemaker was the first to observe patterns of oxygen delivery and cardiac index that were higher in survivors than in nonsurvivors among high-risk surgical cases. Taking this a step farther, Shoemaker et al2o and Boyd et al21 targeted these end points of oxygen delivery, oxygen consumption, and cardiac output and found that among high-risk surgical patients in whom specific values for cardiac index, oxygen delivery, and oxygen consumption were achieved, mortality was reduced from 33% to 4%, a statistically significant decrease in mortality. Other studies involving critically ill patients have found no benefit to achieving specific end points for oxygen delivery. It had no direct effect on mortality rates. Thus, the use of oxygen delivery as a therapeutic parameter for resuscitation is controversial. Gastrointestinal tonometry. Stomach mucosa is used to assess tissue perfusion. If tissue perfusion is adequate to the stomach, as measured by a gastric tonometer, then tissue perfusion must be adequate to tissues that are high on the perfusion ladder. Tonometry is relatively noninvasive. A tonometer is a nasogastric tube with a fluid-filled balloon distally. It measures the partial pressure of CO 2 in the gastrointestinal mucosa by allowing

THE MULTIPLE-INJURYPATIENT equilibration of the partial pressure of CO 2 in the fluid-filled balloon with that in the gastric mucosa. Assuming that excess production of CO 2 occurs during hypoxia, increased CO 2 levels should reflect tissue hypoperfusion. Because the gut is highly susceptible to hypoperfusion, changes in pH may be an early indicator of concurrent global hypoperfusion and impending shock. Many studies support a relationship between reduced gastrointestinal pH and increased patient mortality. 22 Other investigators have not found that gastric pH correlated with survival or nonsurvival. 23 Gastric pH was found not to be a useful end point for resuscitation of these patients. Stratification of patients into subgroups may explain the discrepancy in results. Gastric tonometry is somewhat labor intensive and has strict parameters to ensure reliable measurements. Manufacturers have developed automated devices for sampling of the gastric pH, making it much more clinically friendly. Limitations still include introduction of air into the fluid sample, inaccurate timing while awaiting equilibration of the sample and the gastric fluid, erroneous blood gas analyzer measurements of Pco 2, and high gastric luminal acidity. All clinical and laboratory markers of hypovolemic shock and adequacy of resuscitation have utility and limitations as tools to help in the treatment of patients. In the very dynamic situation of patient resuscitation, no one marker can achieve the goal of indicating that a patient is in this much shock and needs this much resuscitation in this much time, or these organs are going to die and take the entire organism with them. The best we can do is use a combination of markers and look for patterns that support our observations that a patient is getting better or is not getting better. Future methods may encompass a multimodal approach using an integration of cardiac, respiratory, and tissue perfusion variables to assess adequacy of resuscitation. The ability to find a local tissue bed that can represent global tissue perfusion and ensure that it is adequately resuscitated is found in another method; perhaps monitoring of transcutaneous oxygen and carbon dioxide in cutaneous tissue is one such method. The simplicity of the concept of getting oxygen to the tissues is contrasted to the multiple paradigms of trying to ensure this occurs at the tissue level. Treatment to enhance tissue perfusion fall

187 into broad categories that involve manipulation of cardiac output, adequate oxygenation of blood by controlling ventilation, and maintaining intravascular volume with fluids administered intravenously.

RESUSCITATION The entire focus of this article is ability to evaluate organ systems that are injured and maintain tissue perfusion to those tissues. A cornerstone of this is replacement fluids. The body needs adequate fluid tissue to transport oxygen to its member cells. Oxygen gets to the mitochondria by two methods. The first is by mass transport in hemoglobin tissue to the capillary beds, and the second is by diffusion down a concentration gradient across the physical cellular structures. Oxygen diffuses from the blood to the cell by virtue of the pressure gradient between the blood and the cell. The rate at which blood flows to the cells with its cargo of oxygen is just as important as the diffusion pressure of its oxygen load. As blood passes by a cell, the Po 2 difference between the blood and the cell determines diffusion. The higher the blood Po 2, the greater the gradient. Tissues never know how much oxygen is out there, but only if they are receiving enough for aerobic function. Because oxygen is carried to cells by diffusion from a stream flowing by, oxygen reaching the cells is dependent on both oxygen pressure and blood flow. Under resting conditions, we can calculate how much oxygen is transported and how much oxygen diffuses out of the hemoglobin tissue into other tissues for consumption. In a 70-kg man, delivery of oxygen {dot Do2} is the product of hemoglobin concentration (15 g) multiplied by 1.38 (1 g of hemoglobin can hold 1.38 mL of oxygen) multiplied by cardiac output (estimated at 5 L/min) times arterial saturation. Because grams of hemoglobin is expressed in grams per deciliter and cardiac output is expressed liters per minute, the equation must be multiplied by 10: {dot DOE} = (15 X 1.38) x 5.0 • 10 ~- 1,000 mL/min. Normal oxygen consumption {dot VO2} is approximately 250 mL/min. Subtracting consumption from delivery shows a reserve of about 750 cm3/min under resting conditions. A trained athlete can increase cardiac output to 50 L/rain but still experience lactic acid, causing stiff muscles. The cellular con-

188 sumption of oxygen can exceed delivery. As noted earlier, an oxygen debt is created. A sensor that consumes oxygen at the same rate or faster than tissues at risk for hypoxia would facilitate our resuscitation efforts. This sensor could determine whether the oxygen shortage is caused by decreased tension or decreased flow, or how much of each. If flow were the problem, non-red blood cell fluids could be administered. If oxygen tension were the problem, oxygen-carrying solutions, eg, red blood cells, could be administered. Unfortunately, no such sensor exists. We use our clinical judgment based on the previous markers of hypovolemic shock to administer fluids in the hope of ensuring adequate oxygen delivery.

Fluid Therapy Much passion is involved in any discussion of appropriate fluid replacement. Successful resuscitation has been accomplished with every intravenous fluid available. A general overview of the characteristics of different fluids, their associated risks, and their effects on hemodynamic parameters follows. Ultimately, it is up to the clinician to decide the most advantageous use of each fluid. Crystalloids. Crystalloid solutions are water with cations and anions, with or without glucose, in various concentrations and osmolarities to mimic the water and salt milieu of the human body. They are the least expensive of the various solutions and are not allergenic, irnmunogenic, or toxic. They equilibrate across all solute compartments quickly and have an intravascular half-life of approximately 15 minutes. Three times the amount of crystalloid is given for the amount of blood lost. These are usually the first fluids administered for resuscitation in acute hypovolemic shock. They rapidly restore volume and urinary output and keep blood viscosity low. Colloids. Blood product colloids are considered in the next section. Colloids that are used in clinical practice contain a large macromolecular moiety in isoosmotic saline. The common moieties are albumin, polypeptide gelatin, dextran, and hydroxyethylstarch. The colloid moiety has greater water-binding capacity than the Na cation in crystalloid solutions and can retain a large fraction of infused fluid in the vascular space. The increase in plasma volume persists for longer periods with colloid infusion. The blood loss volume is replaced in a one-to-one ratio with colloid solutions. Aller-

MICHAEL J. SULLIVAN genic, coagulopathic, or immunogenic reactions can occur. The synthetic colloids and processed albumin and protein fractions have minimal if any risk of infection. Blood. Volume loss can be corrected with either crystalloids or colloids, but defects in either oxygen-carrying capacity or coagulation can be corrected only by blood and blood products. Current practice is a one-to-one ratio of replacement for each volume of blood lost. The benefit of transfusion is restoration of intravascular volume and oxygen-carrying capacity. The risks of transfusion range from allergic reactions to fatal hemolytic transfusion reactions. Infectious risks range from hepatitis B to human T-cell leukemia lymphoma virus I and II. Massive transfusion. Besides infectious complications, complications unique to massive transfusions include coagulopathy, hypothermia, and metabolic derangements. Massive transfusion is the replacement of more than one blood volume within several hours. Coagulopathy can occur when replacement fluids (crystalloids, colloids, or packed red blood cells) dilute platelets or protein factors to below a level at which they can function. Clinically this is recognized by microvascular bleeding, oozing, or lack of clot formation at wound, surgical, or invasive catheter puncture sites. No single laboratory coagulation test will give complete information on hemostastic function during massive transfusion. Prothrombin time (PT), partial thromboplastin time (PTT), and thrombin time have not been reliable in prediction of perioperative bleeding. Dilutional coagulopathy is caused more commonly by a thrombocytopenia than by a coagulation factor deficit. Platelet transfusion is recommended when platelet counts are less than 50 • 109/L and there is clinical evidence of bleeding. If there is any evidence of prior platelet dysfunction, platelets are administered at intermediate platelet counts of 50 to 100 • I09/L. Other authors believe intraoperative bleeding occurs at platelet counts below 100 • 109/L, which occur after transfusion of approximately 10 units of blood. Both schools believe that platelets should generally be given before fresh-frozen plasma (FFP). For normal hemostasis to occur, approximately 20% of factor V and 30% of factor VIII must be present. Lower levels usually do not occur until

THE MULTIPLE-INJURYPATIENT more than one blood volume has been replaced. In the clinical setting, when PT and PTT cannot be obtained in a timely fashion, if there is evidence of bleeding and with replacement of one blood volume, FFP can be administered. The risks of a transfusion reaction are equal for FFP and red blood cells. Two units generally will correct most coagulopathies. Metabolic problems associated with massive transfusions are citrate toxicity, hypocalcemia, hyperkalemia, hypothermia, and acidosis. Citrate toxicity is secondary to the administration of large quantities of citrated blood components. This can acutely lower the serum calcium level. The liver quickly metabolizes citrate; however, in states of shock, liver function is impaired and citrate may not be cleared as rapidly. Citrate induces hypocalcemia, causing hypotension, narrowing of the pulse pressure, increased cardiac filling pressures, gross muscle tremors, and prolonged QT interval on electrocardiogram. All of these signs except for prolonged QT interval are seen in shock, regardless of calcium level, and therefore are not specific for hypocalcemia in the situation in which it is most likely to be encountered, massive transfusion. Exogenous calcium is given when the measured ionized calcium level is low or falling in the face of ongoing blood transfusions. Hyperkalemia is seen with massive transfusions that require blood replacement in a short time. The older the unit of red blood cells, the higher the potassium level in the stored blood. Potassium leaves viable erthrocytes and increases the concentration in plasma of stored blood. After 21 days of storage, plasma potassium concentrations approach 25 to 30 mEq/L. Massive transfusions that occur over 8 hours may never necessitate treatment for hyperkalemia because the potassium can redistribute to the intracellular space. Current rapid-infusion technology allows blood infusion of 100 mL/ min per machine. If two rapid-transfusion devises are being used simultaneously, acute hyperkalemia can occur. The best clinically useful intraoperative marker for acute hyperkalemia is the electrocardiogram. Peaking of the T waves signifies a potassium level that is clinically significant. Laboratory values of potassium may not be significantly high, but the associated hypocalcemia narrows the range of safety for potassium levels. Treatment of hyperkalemia ranges from surgical control of hemorrhage; administration of calcium; hyperventilation; ad-

189 ministration of bicarbonate, insulin, and glucose; and, rarely, administration of epinephrine. Hypothermia occurs in massive transfusions for multiple reasons: environmental exposure before arrival, low ambient operating room temperatures, large exposed wound surfaces, anesthetic drugs, and infusion of fluids below body temperature. Warming of blood before infusion is essential to combat intraoperative hypothemia. Infusing stored blood is infusing an acid load. The pH of stored blood is 6.6 to 6.9 because of the slow accumulation of carbon dioxide and lactic acid from erythrocyte metabolism and the citric acid in the anticoagulant. With normal tissue perfusion, this acid load is rapidly metabolized and lactate and citrate are converted to bicarbonate in the liver. This alkalinization is variable, depending on the functional status of the liver, and is rarely a problem in massive transfusions. Options in emergent transfusions. When transfusion of red blood cells is emergent, what are the options? The universal donor blood is type O Rhnegative red blood cells. This can be given immediately without laboratory verification of compatibility. If two units or less of this type of blood is administered, the patient's own blood type still can be given subsequently. When larger amounts of universal donor blood are administered, transfusion with the patient's own typed and crossedmatched blood can lead to transfusion reactions. The second option is type-specific red blood cells. Typing considers ABO and Rh blood antigens. ABO- and Rh-compatible blood can be selected from the stock and issued within 5 to 10 minutes. Delays in receiving type-specific blood are usually caused by transportation of specimens and red blood cells between the laboratory and the patient. Type-specific blood is safe and has a very low potential for a transfusion reaction. Those at higher risk include patients who have previously received transfusions and multiparious woman, who may harbor undetected antibodies. A third option is administration of type O Rhpositive red blood cells when blood must be administered immediately. Women who are past child-beating age or have undergone a hysterectomy and men can receive Rh-positive blood. This increases the pool of universal donor blood available for transfusion. This blood should be considered when the above two options are not available. For typed and crossed-matched blood, at least 45

190

M I C H A E L J. SULLIVAN

minutes are needed to check compatibility. The patient's serum is incubated in a test tube with an aliquot of red blood cells from a specific donor unit to verify in vitro compatibility. Both ABO and all other red cell antigens can be tested. In acute hypovolemic shock situations, intraoperative blood cell salvage can be used to decrease the amount of blood bank blood transfused. With massive blood loss, both salvaged and typed blood may be needed. The decision to transfuse blood should not be based on an arbitrary hemoglobin threshold below which transfusion is begun. A young patient with a single episode of blood loss who is currently not bleeding and has a hemoglobin of 8 may not need red blood cell transfusion, whereas a geriatric patient with duodenal bleeding and a hemoglobin of 10 may require red blood cells at the time of surgical incision. The decision depends on the nature of the injuries, underlying physiological reserves, general health, and the rate at which blood is being lost or is expected to be lost.

Hymodynamic changes of various fluids. When fluids are administered for resuscitation in critically injured patients, what effects do the various fluids have on hemodynamic and oxygentransport parameters? Shoemaker and Wo, z4 using invasive and noninvasive methods, examined this question in various patient populations from the emergency department to the operating room to the intensive care unit (Table 2). If the best is given a rank of 1 and the worst is given a rank of 4, and this is done for each variable, fluids can be ranked from first to fourth in

terms of the most to the least effective in improving overall hemodynamic status. The fluid order, from most to least effective, is whole blood, albumin, packed red blood cells, and lactated Ringer's solutions. Understanding the effects of the various fluids on hemodynamic parameters can lead to greater precision in fluid administration. POSTOPERATIVE C O M P L I C A T I O N S

Successful, appropriate intraoperative resuscitation is not a guarantee for an unremarkable postoperative course. Early aggressive resuscitation can lessen the chance of postoperative organ failure. Mounting evidence shows that a hypoperfused, hypoxic, and acidotic tissue environment leads to regional and remote-site organ failure. Multiple animal studies of shock and resuscitation show that after a hemorrhagic episode followed by resuscitation to normal blood pressure, capillary blood flow in various tissues several hours later is only 20% to 40% of normal. 25 Thus, although patients appear to be resuscitated intraoperatively, ongoing hypoxia at the tissue level may not be reversed. This manifests itself as organ failure or dysfunction in the intensive care unit after surgery; the primary organs are the lungs and the kidneys. The current standard of care is fluid replacement with various combinations of crystalloid and colloid solutions until restoration of pressure is deemed adequate and various laboratory markers are normalized. Currently, at the University of Southern California Medical Center, we are studying a resuscitation protocol that actively opens the microcirculation while concurrently restoring vol-

Table 2. Hemodynamic and Oxygen Transport Changes After Various Fluid Therapies in ICU Patients by Invasive Monitoring

Variable MAP (mm Hg) CI {L/min/m 2) POAP (mm Hg) Do 2 (mL/min/m 2) Vo2, (mL/min/m 2)

WB (1 U) n = 86 8.8 0.44 2.7 87 24

_ 1.3" _+ 0.09* • 0.9 t + 14" + 6*

PRBC (1 U) n = 32 3.6 0.02 1.3 64 7

_ 1.9t +_ 0.4 • 0.8 _+ 22* +_ 7

Albumin (500 mL) n = 82 4.7 0.67 4.0 65 10

_-_ 1" _ 0.09* - 1.4" - 14" _ 4t

LR (1,000 mL) n = 35 5.9 0.23 2.3 -3 0

_+ 2.1' _ 0.14 • 1.7 • 25 • 7

NOTE. Data are expressed as means + SEM. Abbreviations: ICU, intensive care unit; WB, whole blood; PRBC, packed red blood cells; LR, lactated Ringer's solution; MAP, mean arterial pressure; CI, cardiac index; PAOP, pulmonary artery occlusion pressure. Do2, oxygen delivery; V02, oxygen consumption. * P < .01. t p < .05%.

THE MULTIPLE-INJURYPATIENT ume. Nitroglycerin is used as part of the fluid restoration while mean arterial pressure is kept at a safe level. Escalating nitroglycerin doses are used in an attempt to open neuroendocrine-mediated vasoconstricted capillaries, specifically venousside capillaries. This has favorable effects on capillary hemodynamics compared with the detrimental effects of arterial vasodilators. Vasodilators have been used in animal models of shock since the 1940s but have fallen out of favor. With new understanding of the pathophysiology of the hemorrhagic shock state and the role of nitric oxide as an endothelial vasodilator, perhaps better resuscitation protocols will be developed. The theory is that the more quickly flow is established to the microcirculation, the less organ failure and dysfunction there will be in the postoperative period secondary to a hypoxic, acidotic environment.

REFERENCES 1. Starling E: On the absorption of fluids from the connective tissue spaces. J Physiol (Lond) 19:312-316, 1896 2. Wiederhielm CA: Dynamics of transcapillary fluid exchange. J Gen Physiol 52:29s-63s, 1968 3. Guyton AC: Textbook of Medical Physiology. Philadelphia, PA, Saunders, 1981, pp 183-185 4. Taylor AE: Capillary fluid filtration, Starling forces and lymph flow. Circ Res 49:557-575, 1981 5. Rose BD: Physiology of body fluids, in Rose BD (ed): Clinical Physiology of Acid-Base and Electrolyte Disorders (2nd ed). New York, NY, McGraw-Hill, 1984, pp. 3-42 6. McLeod MK, Carlson DE, Gann DS: Hormonal responses associated with early hyperglycemia after graded hemorrhage in dogs. Am J Phyiol 251:597-603, 1986 7. Davidson JS, Birdsell DC: Cervical spine injury in patients with facial skeletal trauma. Trauma 29:1276-1278, 1989 8. ACC/AHA Task Force Report: Guidelines for perioperative cardiovascular evaluation for noncardiac surgery: Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation 93:1278-1317, 1996 9. Chun K, Zhang J, Biewer J, et al: Microcirculatory failure determines lethal hepatocyte injury in ischemic-reperfused rat livers. Shock 1:3-9, 1994 10. Runciman WB, Skowronski GA: Pathophysiology of hemorrhagic shock. Anaesth Intensive Care 12:193-205, 1984

191 11. Wang P, Ba ZF, Burkhardt J, et al: Trauma--Hemorrhage and resuscitation in the mouse: Effects on cardiac output and organ blood flow. Am J Physiol 264:1166-1173, 1993 12. Bellamy RF, Pederson DC, DeGuz~an LR: Organ blood flow and the cause of death following massive hemorrhage. Circ Shock 14:113-127, 1984 13. Reed MK, Taylor BJ, Smith G, et al: Splanchnic prostanoid production: effect of hemorrhagic shock. J Surg Res 48: 579-583, 1990 14. Flynn WJ, Cryer HG, Garrison RN: Pentoxifylline restores intestinal microvascular blood flow during resuscitated hemorrhagic shock. Surgery 110:350-356, 1991 15. Scannell G, Clark L, Waxman K: Regional flow during experimental hemorrhage and crystalloid resuscitation: persistence of low flow to the splanchnic organs. Resuscitation 23: 217-225, 1992 16. Wang P, Hauptman JG, Chaudry IH: Hemorrhage produces depression in microvascular blood flow which persists despite fluid resuscitation. Circ Shock 32:307-318, 1990 17. Perbeck L, Lund F, Thulin L: Intestinal capillary blood flow studies with fluorescein flowmetry in hemorrhagically shocked rats. Acta Chir Scand 151:657-661, 1985 18. Davis JW, Shackford SR, Mackersie RC, et al: Base deficit as a guide to volume resuscitation. J Trauma 28:14641467, 1990 19. Abramson D, Scalea TM: Lactate clearance and survival following injury. J Trauma 35:584-589, 1993 20. Shoemaker WC, Appel PL, Kram HB, et al: Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients. Chest 94:1176-1186, 1988 21. Boyd O, Grounds RM, Bennet ED: A randomized clinical trial of the effect of deliberate perioperative increase of oxygen delivery on mortality on high-risk surgical patients JAMA 270:2699-2707, 1993 22. Friedman G, Berlot G, Kahn RJ, et al: Combined measurement of blood lactate concentrations and gastric intramucosal pH in patients with severe sepsis. Crit Care Med 23:1184-1193, 1995 23. Joynt GM, Lipman J, Gomersall CD, et al: Gastric intramucosal pH and blood lactate in severe sepsis. Anaesthesia 52:726-732, 1997 24. Shoemaker WC, Wo CCJ: Circulatory effects of whole blood, packed red cells, albumin, starch and crystalloids in resuscitation of shock and acute critical illness Vox Sanguinis 74:69-74, 1998 (suppl 2) 25. Zhao KS, Junker D, Ledano FA, et al: Microvascular adjustments during irreversible hemorrhagic shock in rat skeletal muscle. Microvasc Res 30:143-153, 1985 26. Klein HG: Allogenic transfusion risks in the surgical patient. Am J Surg 170:21-26, 1995 (suppl 6A)