Brian L. Erstad
Plasma does not have to be lost from the body for hypovolemic shock to occur.
Patients may die of hypovolemic shock despite having normal serum electrolyte concentrations.
Although the Starling’s equation of fluid transport is useful for understanding the factors involved in fluid shifting between compartments, it is not a practical tool for use in the clinical setting.
Patients may have complications and death as a result of reperfusion injury as well as the initial insult.
The clinical presentation of patients with hypovolemic shock can vary substantially, depending on concomitant disease states, medications, and cause of hypovolemia.
The initial monitoring of a patient with suspected intravascular depletion always should include vital signs, urine output, mental status, and physical examination.
The need for IV (vs. oral) rehydration in children often is overestimated.
Crystalloid (sodium-containing) solutions should be used for most forms of circulatory insufficiency that are associated with hemodynamic instability.
Neither crystalloids nor colloids have the oxygen-carrying properties of red blood cells.
Vasoactive medications should not be considered for hypovolemic shock until fluid resuscitation has been optimized.
This chapter discusses the assessment and management of hypovolemic shock. Neurogenic shock resulting from loss of sympathetic activity and anaphylactic shock resulting from increased vascular permeability often are considered separately from hypovolemic shock because fluid loss from the body is not necessary for their occurrence. Although these forms of shock are not discussed in detail, it is important to note that IV fluid administration (in conjunction with vasoactive medications) is a mainstay of therapy because circulating volume is decreased. In this regard, adequate fluid resuscitation to maintain circulating blood volume is a common principle in managing all forms of shock.
Because shock is not a reportable category by state and federal agencies that track causes of death, the incidence is unknown. Estimates of deaths due to shock are complicated by differences in definitions and classification systems. Part of the problem is defining when progressive circulatory insufficiency results in the loss of normal compensatory responses by the body, which could reverse the processes leading to irreversible organ dysfunction. This loss of appropriate compensation varies from patient to patient and is not always readily apparent during the initial patient presentation. Therefore, forms of hypovolemic shock, such as hemorrhagic shock, are subsumed by more readily identifiable categories of death, such as accidental injuries and homicides. Crude and conservative estimates of death due to hypovolemic shock are available for some of its forms. More than 100,000 deaths each year in the United States are due to unintentional injuries that frequently involve bleeding,1 and more than 600 deaths each year are due to natural heat-related illness.2 The figures are much higher when considered on a global basis. For example, electrolyte depletion and dehydration due to diarrheal disease result in approximately 2 million deaths each year in children younger than 5 years.3 The most liberal estimates of death include all causes of circulatory failure (i.e., the last stage of shock).
Hypovolemic shock may result from blood loss (plasma and red blood cells) due to trauma, surgery, or internal hemorrhage or from plasma loss due to fluid sequestered within the body or lost from the body (Table 14-1). In some cases, such as in postoperative patients, a number of these problems occur at the same time. For example, a patient may have blood loss secondary to trauma or surgery, with additional fluid being third spaced (e.g., as tissue edema in the GI tract with a concomitant ileus) and lost through a high-output GI fistula postoperatively. As this example of third-spaced fluid indicates, fluid (i.e., plasma) does not have to be lost from the body for a person to develop hypovolemic shock, although the fistula output would clearly aggravate the situation. Approximately 10 L of fluid is secreted and reabsorbed daily in the GI tract; so, it is not surprising that volume loss could be substantial depending on the location of the fistula and function of the tract preceding the fistula.
TABLE 14-1 Causes of Hypovolemic Shocka
Dehydration may result from primary water deficiency, usually because of decreased intake, but in some instances (e.g., diabetes insipidus) it may result from increased losses of water. With most forms of dehydration, such as those caused by diarrheal disease and heat-related illness, a combination of inadequate intake and higher than normal losses occurs. In general, the term dehydration implies primary intracellular water depletion, in contrast to volume depletion, which implies extracellular, and particularly intravascular, sodium and water loss. However, there is substantial overlap in the definitions and use of terms such as dehydration and volume depletion in the medical literature, so the reader must be cognizant of the intended meaning. In the case of primary water deficit, cell dehydration occurs. Initially, the patient may be thirsty and possibly have some mental status changes, such as confusion. If cellular dehydration occurs slowly, intracellular substances, referred to as idiogenic osmoles, develop that limit progressive complications (e.g., cerebral edema or coma). Death due to primary water deficit, if it occurs, is usually a result of delayed circulatory failure. With combined water and salt deficiencies, such as might occur with GI (e.g., diarrhea) and skin losses (e.g., heat stroke), interstitial and intravascular depletion is an early occurrence. Fortunately, dehydration is relatively easy to prevent with routine vigilance and water replacement compared with some of the other causes of shock.
Hypovolemic shock often is described in terms of monitoring parameters such as lowered blood pressure, but patients with shock may die despite normal surrogate markers of circulatory insufficiency. Therefore, an appropriate definition should mention the underlying problem, which is inadequate tissue perfusion resulting from circulatory failure. In the case of hypovolemic shock, the cause of the altered perfusion is fluid (or volume) depletion resulting from trauma, surgery, thermal injury, or some form of dehydration. Figure 14-1 provides a simplified view of the pathophysiology of circulatory insufficiency assuming the acute insult causing the plasma volume depletion did not result in immediate patient death. Cell damage and death may occur from the primary insult or from reperfusion injury. The latter problem is associated most frequently with trauma and blood loss that cause a systemic inflammatory response syndrome (SIRS) with the release of a multitude of mediators of inflammation and injury that have complex interactions. Cells have varying responses to hypoxia, ranging from astrocytes that quit functioning almost immediately to other cells that may tolerate more prolonged periods of hypoperfusion. Left unmitigated, cell death occurs with prolonged injury and is usually heralded by acidosis, hypothermia, and coagulopathy—referred to as the lethal triad.
FIGURE 14-1 Pathophysiology of circulatory insufficiency and failure (shock).
The body attempts to compensate for volume depletion beginning with autoregulatory changes involving smaller blood vessels. When the cause of circulatory insufficiency continues unabated, local mechanisms eventually fail to provide adequate compensation, and macrocirculatory changes ensue. The majority of blood volume is contained in venous capacitance vessels, with gravity being the major impedance to flow back to the heart. With increasing volume depletion, blood flow to the heart (preload) is decreased, with subsequent activation of baroreceptors and chemoreceptors leading to sympathetic discharge. Also, fluid shifting from the interstitial space to the intravascular space occurs through a phenomenon known as transcapillary refill, and hormones (e.g., adrenocorticotropic hormone, angiotensin, catecholamines, and vasopressin) that cause sodium and water retention by the kidneys are released. The phenomenon of transcapillary refill means that the body can have fluid losses exceeding normal plasma volume. These responses cause alterations in stroke volume, heart rate, and peripheral vascular resistance so that blood pressure and hence tissue perfusion can be maintained.
The microcirculatory changes associated with shock are complex and difficult to study. Although some mediators such as catecholamines, angiotensin II, arginine vasopressin, and endothelin-1 cause vasoconstriction, other mediators, such as adenosine and nitric oxide, yield vasodilation. These changes result in hypoperfusion or hyperperfusion, depending on the organs involved. As these microcirculatory changes fail to maintain adequate organ perfusion, more widespread sympathetic nervous system activation and vasoconstriction ensue. Even assuming general circulation is restored, capillaries may not function properly due to ongoing edema and ischemia. Failure to respond to sympathetic stimulation and fluid administration is indicative of the vasodilation that occurs in the final phase of circulatory failure leading to death.
The factors involved in fluid shifting between the intravascular and interstitial spaces are described by the modified Starling’s equation:
where Jv is the net transvascular flow rate (cannot be measured in the clinical setting), Kf,c is the capillary filtration coefficient for fluids (cannot be measured in the clinical setting), Pc is the capillary hydrostatic pressure (indirectly estimated in the clinical setting, e.g., pulmonary artery occlusive pressure), Pt is the tissue hydrostatic pressure (cannot be measured in the clinical setting), σ is the reflection coefficient for proteins (cannot be measured in the clinical setting), πc is the plasma colloid osmotic pressure (not usually measured in the clinical setting, but technology is available), and πt is the tissue colloid osmotic pressure (cannot be measured in the clinical setting). The physiology of the parameters associated with the Starling equation continue to be elucidated. For example, the plasma colloid osmotic pressure is more appropriately referred to as the oncotic pressure within the endothelial surface layer and the tissue colloid osmotic pressure is more appropriately referred to as the oncotic pressure beneath the endothelial surface layer.
Proteins act as oncotic agents in each of these spaces to attract fluid, whereas hydrostatic forces push fluid into or out of the vessels. The equation has distinct permeability values for water and protein because each crosses the vascular membrane at a different rate. The values for the variables listed in the equation are not the same for capillaries in all parts of the body. For example, on a scale from 0 to 1 with 0 being free passage of protein and 1 being impermeable to protein, the typical value for the reflection coefficient in most capillaries is >0.9. However, in the pulmonary capillaries the value is closer to 0.7 and approaches 0 in inflammatory states associated with increased capillary permeability.4 As the value approaches 0, the capillaries are freely permeable not only to the usual fluid and electrolytes but also to plasma proteins such as albumin. Because albumin accounts for approximately 80% of the plasma oncotic pressure, its free passage into the interstitial space effectively negates its intravascular oncotic benefit. Although the Starling’s equation is useful to practitioners in terms of understanding the factors involved in fluid shifting between compartments, the rate and direction of transvascular flow cannot be calculated accurately in the clinical setting because most factors cannot be measured directly and the values for the factors vary in different capillaries in the body.
The body’s compensatory mechanisms may have beneficial and harmful consequences. For example, cardiac output can be increased substantially by increases in stroke volume or heart rate. Although this may be useful for providing blood flow to inadequately perfused tissues, it may cause large increases in oxygen consumption by the heart that could aggravate preexisting ischemia in patients with underlying coronary artery disease (CAD). Another example is the sympathetic nervous system–mediated vasoconstriction that causes blood to shift from the skin, skeletal muscle, and some internal organs such as the kidneys and GI tract to organs (e.g., heart and brain) that are less tolerant of inadequate flow. If the vasoconstriction continues unabated, the hypoperfused organs eventually become damaged. Figure 14-2provides an overview of the compensatory changes that occur with a loss of circulating blood volume.
FIGURE 14-2 Activation of compensatory mechanisms with loss of circulatory volume. Certain stages may be absent, depending on a number of factors, such as age, preexisting disease states, and cause of circulatory insufficiency. (ACTH, adrenocorticotropin; ANP, atrial natriuretic peptide; BP, blood pressure; CO, cardiac output; HR, heart rate; PVR, peripheral vascular resistance; RR, respiratory rate.)
In addition to the more acute implications of hypovolemia and attendant complications, reperfusion damage is likely to occur, particularly after prolonged resuscitation attempts. In addition to edematous obstruction of capillaries and oxygen-free radical damage of cell membranes, a number of cellular (e.g., white blood cells and platelets) and humoral (e.g., procoagulants, anticoagulants, complement, and kinins) components are activated, causing the release of other inflammatory mediators. The resulting reperfusion injury may range from readily reversible organ dysfunction to multiple-organ failure and death. The lungs are frequently the first system affected either by excessive fluid resuscitation or by the mediators of secondary reperfusion injury. The latter form of injury often results in the acute respiratory distress syndrome that is defined by an arterial oxygen tension-to-fraction of inspired oxygen ratio of less than or equal to 300 (with additional subdivisions of mild, moderate, and severe) with bilateral lung opacities in the absence of hypervolemia.
Although the basic pathophysiology is similar for the various causes of hypovolemic shock, there are unique considerations relative to each. For example, whereas isolated head injuries associated with trauma typically do not result in substantial blood loss or shock, long bone or pelvic fractures may sequester several liters of blood. Patients with traumatic or thermal injuries, as well as postoperative patients, may have substantial fluid accumulation in sites where the fluid cannot be readily transferred back into blood vessels (i.e., third-spaced fluid) for maintaining pressure. With these types of injuries, prompt control of compressible bleeding sources with rapid patient transfer to the hospital for definitive treatment may preclude the cascade of events leading to shock. Indeed, with trauma patients, a “scoop and run” approach that places a priority on rapid transport to a hospital is used by most urban hospitals.
In the case of hemorrhagic shock, prompt attention must be given to cell as well as plasma losses. Red blood cells lost during the bleeding episode may lead to ischemic damage in vital organs. Packed red blood cell transfusions may be needed to increase the oxygen-carrying capacity of the blood because oxygen transport is a function not only of cardiac output but also of hemoglobin concentration and saturation and of hemoglobin affinity for oxygen. Once hemostasis has been achieved, a more restrictive transfusion strategy (i.e., transfusion if hemoglobin <7 g/dL [<70 g/L; 4.34 mmol/L]) is indicated for the majority of patients without severe cardiovascular disease (see Trauma/Perioperative Patients below).
Clotting factors and platelets are also lost in hemorrhage. The resulting bleeding problems may be aggravated by the dilutional effect of fluid resuscitation on clotting factor activity. Fresh-frozen plasma that contains necessary clotting factors and platelets is needed in massive blood loss to restore adequate coagulation. On the other hand, trauma patients are at increased risk for deep vein thrombosis and pulmonary embolism caused by multiple factors, including vessel damage, abnormal blood flow patterns, and the hypercoagulable state associated with injury. Therefore, some form of venous thromboembolism prophylaxis usually is indicated in multiple-trauma patients or patients with severe single-system injuries (e.g., spinal cord damage) once hemostasis of major injury-related bleeding has been achieved.
The pathophysiology becomes more complicated if the severity of shock is sufficient to require patient admission to the intensive care unit (ICU) after initial resuscitation or surgery. Most patients admitted to the ICU have SIRS, which is the body’s response to injury. This syndrome is defined by a number of hypermetabolic changes reflected in the patient’s temperature, white blood cell count and differential, and respiratory and heart rates. The stress response involves complex interactions between the nervous system and immunomodulating substances and has similar (if not the same) harmful and helpful consequences described with reperfusion following shock. If the underlying problems are left untreated, the patient with SIRS may develop multiple-organ dysfunction syndrome (MODS) during the final stages of illness.
The initial presentation of patients with suspected volume depletion can vary markedly, depending on factors such as age, concomitant disease states and medications, and the etiology and rapidity of depletion (see Clinical Presentation of Hypovolemic Shock box). Intravascular depletion as a consequence of blood loss is signified by postural vital sign changes (i.e., changes in pulse and blood pressure between supine, sitting, and standing measurements), and such measurements should be performed unless the diagnosis is obvious, as in the case of bleeding associated with trauma. Early signs and symptoms of dehydration and intravascular depletion caused by GI or urinary losses often are relatively nonspecific. Plasma volume losses of <10 mL/kg of body weight usually are associated with minor signs and symptoms of distress. Larger losses are not likely to be well tolerated (Table 14-2), particularly in patients older than 65 years. An 18-year-old athlete and a 65-year-old sedentary individual are likely to have much different responses to a similar amount of fluid loss. The young patient may lose one-fourth of his or her circulating blood volume with minimal changes in arterial blood pressure and a relatively low heart rate. However, the elderly patient may have orthostatic changes in blood pressure that are not well tolerated by organs such as the kidneys. Unfortunately, this same elderly patient may not have common signs and symptoms of volume depletion, such as skin turgor changes or thirst, but instead may have more subtle changes (e.g., mental status alterations).
TABLE 14-2 Acute Circulatory Insufficiency: Initial Presentation and Therapya
The diagnosis of dehydration and intravascular depletion in children is complicated by difficulties in obtaining an accurate history. However, some excellent resources are available for healthcare providers, such as the Centers for Disease Control and Prevention (CDC) guidelines (www.cdc.gov), which discuss the evaluation and management of diarrhea in patients of all ages. In younger children, parental observations are important for estimating fluid deficits and deciding whether hospitalization is necessary. Fortunately, prospective data suggest that parental histories are predictive of acidosis and the need for hospitalization.5 Regardless of patient age or preexisting conditions, the initial monitoring of a patient with suspected volume depletion should include the following noninvasive parameters: vital signs, urine output, mental status, and physical examination (Fig. 14-3). An increase in blood pressure with passive leg raising may also be useful for the assessment of suspected hypovolemia, but should not be used to guide responsiveness to fluid administration.
FIGURE 14-3 Noninvasive assessment of circulatory insufficiency.
Although the presenting signs and symptoms of circulatory insufficiency are variable, patients usually have decreased blood pressure, increased heart and respiratory rates, and a normal or low–normal temperature (e.g., 36°C to 37°C [96.8°F to 98.6°F]) in the absence of infection, exposure to extremes of temperature, and medications that impair thermoregulation. As mentioned earlier, recordings of vital signs must be interpreted in light of known or suspected baseline conditions. For example, alcohol, β-blockers, diuretics, and medications with anticholinergic effects may impair thermoregulation. Medications such as β-blockers and calcium channel blockers may alter resting blood pressure and heart rate, as well as the subsequent response to therapeutic interventions.
Although a blood pressure reading of 110/70 mm Hg (systolic/diastolic) may be acceptable in many patients, it may be inadequate in a patient with preexisting hypertension who normally has a blood pressure of 170/105 mm Hg. At the other extreme, patients with very low blood pressure may have inaudible or inaccurate determinations with cuff (sphygmomanometric) measurements. Chapters 1 and 3 detail blood pressure measurement (e.g., cuff size, position). In this case, intraarterial monitoring is indicated. The respiratory rate may be elevated because of anxiety or as a compensatory mechanism for the metabolic acidosis caused by lactic acidosis associated with poor tissue perfusion.
Although the kidneys continually produce urine, the bladder stores the urine for intermittent elimination. For the initial diagnosis and management of acute circulatory insufficiency, a catheter can be inserted into the bladder for measuring urine output. In contrast to thirst, which is a relatively insensitive indicator of volume depletion, urine output is generally diminished with inadequate fluid administration and increases with appropriate resuscitation. This presumes, of course, that acute renal failure or medications such as diuretics are not altering the expected response. Adults should produce at least 0.5 to 1 mL/kg/h of urine, whereas children up to 12 years of age should produce at least 1 mL/kg/h (2 mL/kg/h if younger than 1 year).
Mental status changes associated with volume depletion, if present, may range from subtle fluctuations in mood to unconsciousness. Although the latter finding typically is indicative of more severe depletion, less dramatic findings should not be interpreted as indicating mild fluid deficits. Losses of 3 to 4 L of plasma volume may be associated only with lassitude in an otherwise healthy adult patient. Similar interpretation difficulties must be considered when performing the initial physical examination. An orderly progression from warm, reddish skin with appropriate capillary refill (rapid return of blood flow to the extremity after removal of compression) to cold, cyanotic discoloration with impaired refill may not occur. Also, dry mucous membranes in elderly patients may be caused by mouth breathing or medications and not by fluid depletion.
Milder forms of volume depletion may be managed in outpatient settings. For example, supplemental fluids can be added to the usual estimated daily requirements of 30 to 35 mL/kg in patients older than 12 years with dehydration. Commercially available carbohydrate/electrolyte drinks generally are more palatable than water and may promote earlier recovery. The rationale for combining carbohydrates with sodium is based on the cotransport absorption mechanism in the intestinal tract. With diarrheal states in particular, sodium absorption is impaired. Because water follows sodium, the diarrhea is likely to continue despite oral crystalloid fluid administration until the intestinal pathology resolves. However, when dextrose and sodium are combined in 1:1 equimolar amounts, both are absorbed via the cotransport mechanism, which also allows for absorption of water. This concept forms the basis for the World Health Organization’s (WHO’s) oral rehydration solution, which contains 75 mmol/L of dextrose, 75 mmol/L of sodium, 20 mmol/L of potassium, 65 mmol/L of chloride, and 10 mmol/L of citrate for a total osmolarity of 245 mOsm/L.3 Commercially available nonprescription rehydration drinks for children in the United States also have an osmolarity of approximately 250 mOsm/L but typically contain 50 mmol/L or less of sodium, and the dextrose-to-sodium ratio often is 3:1. How these differences between commercially available formulations and the WHO rehydration formula might affect hospitalization rates is unclear, but ad hoc attempts to alter the commercially available products to make them more consistent with the WHO formula may be dangerous and are not recommended.
CLINICAL PRESENTATION Hypovolemic Shock
• The initial presentation of adult patients with suspected volume depletion could vary markedly, depending on factors such as age, concomitant disease states and medications, and the etiology and rapidity of depletion.
• Plasma volume losses of <10 mL/kg of body weight usually are associated with minor signs and symptoms of distress.
• Patients may present with thirst, nausea, anxiousness, weakness, light-headedness, and dizziness.
• Patients may report scanty urine output and dark yellow urine.
With more severe volume loss:
• Patients would have marked increases in heart rate (e.g., >120 beats/min) and respiratory rate (e.g., >30 breaths/min).
• Blood pressure would be decreased (e.g., systolic blood pressure <90 mm Hg).
• Mental status changes or unconsciousness may occur.
• Agitation may be present if the patient is conscious.
• Body temperature would be low or normal (e.g., 36°C to 37°C [96.8°F to 98.6°F]) in the absence of concomitant infection with cold extremities and decreased capillary refill on physical examination.
• Sodium and chloride concentrations usually are high with acute depletion but may be low or normal depending on type of fluid intake.
• The ratio of blood urea nitrogen (BUN) to creatinine is likely to be elevated initially, but the creatinine level would increase as renal dysfunction occurs.
• Elevated base deficit and lactate concentrations in conjunction with decreased bicarbonate concentrations and pH due to metabolic acidosis.
• The complete blood count should be normal in the absence of concomitant disease states such as infection; in hemorrhagic shock, the red cell count, hemoglobin, and hematocrit would decrease over time, while the prothrombin time (PT) and international normalized ratio would increase.
• With more severe volume depletion, other organs may become dysfunctional, which may be reflected in laboratory testing (e.g., elevated transaminase levels with hepatic dysfunction).
Other Diagnostic Tests
• Urine output would be decreased to <0.5 to 1 mL/h.
Outpatient rehydration of children usually is recommended for those with uncomplicated (e.g., vomitingless than 48 hours) acute gastroenteritis and relatively mild dehydration after the exclusion of more severe illnesses such as bowel obstruction. The need for IV rehydration often is overestimated. Randomized studies conducted in pediatric emergency departments have found oral or nasogastric rehydration to be at least as effective as IV rehydration using end points such as length of stay and need for hospital admission.6,7 While dehydration is primarily a problem of intracellular fluid depletion, ongoing losses will result in extracellular fluid depletion as well. The remainder of this chapter will focus on more severe forms of volume depletion (i.e., hypovolemic shock) that are not amenable to oral rehydration.
The desired outcomes of therapy for patients with hypovolemic shock are to reduce morbidity and mortality by preventing disease progression with subsequent organ damage and, to the extent possible, to reverse organ dysfunction that has already taken place.
Reduce morbidity and mortality by preventing disease progression with subsequent organ damage.
To the extent possible, reverse organ dysfunction that has already taken place.
General Approach to Treatment
Hospitalization is indicated for more severe forms of circulatory insufficiency. If access to the circulatory system for administration of fluids and medication is not obtained prior to hospitalization, this should be a priority. Venous access generally is obtained during the preliminary examination process that includes the ABCs of life support (i.e., airway, breathing, and circulation), assessment of vital signs and mental status, and determination of urine output after catheterization. Whenever large-volume fluid resuscitation is expected, as in hemorrhagic shock, at least two IV catheters are desirable. Because flow is a function of tubing length and catheter diameter, large-bore peripheral IV lines are preferred over longer central lines. Unfortunately, vascular access in some patients may be problematic, and other routes such as intraosseous infusion may be necessary. Prior to the last decade, use of intraosseous fluid and drug administration in the United States was mostly restricted to children with IV access issues, but it is increasingly being used in adult patients as well. One interesting method of fluid administration that has been investigated in elderly patients is subcutaneous infusion, or hypodermoclysis. With hypodermoclysis, common dextrose- and sodium-containing fluids typically given by the IV route are given by subcutaneous infusion at sites such as the upper arm, chest, abdomen, or thigh, depending on factors such as patient or provider preference. Hyaluronidase has been used as a spreading agent to facilitate fluid absorption by this route, but its benefit versus risk profile has yet to be clearly elucidated; in particular, allergic reactions with this agent have been a concern, although a recombinant form is now available that has the potential for fewer reactions compared with the older bovine-derived products. Hypodermoclysis is not used commonly in the United States, probably because of concerns of adverse effects that were found in early studies that used excessively hypotonic or hypertonic solutions, as well as issues related to reimbursement when considered in ambulatory, home, or palliative care settings. Although relatively high fluid administration rates have been achieved in some studies involving hypodermoclysis, this method of infusion should not be used in patients with more severe forms of dehydration or hypovolemia until additional supportive information from clinical trials is available. Although alternative methods of fluid administration, such as hypodermoclysis, are desirable, well-conducted trials are needed before such methods can be recommended for routine use.
After the immediate postresuscitation phase of the treatment of hypovolemic shock, proper attention must be paid to general supportive care measures that include appropriate assessment and management of pain, anxiety/agitation, and delirium. This is particularly true for patients who develop shock after trauma, surgery, or thermal injury and require admission to an ICU.
Nonpharmacologic therapy for shock is dependent on the inciting event, although the basic life support measures such as a secure airway with appropriate oxygenation apply to all patients. For patients with more severe traumatic injury, additional measures would include surgery, stabilization of fractures, control of blood loss by physical compression or surgical control, and prevention of heat loss since hypothermia may aggravate other problems such as bleeding. For patients with heat exposure, cooling measures are indicated. Patients with thermal injuries should have the wound sites covered with cool, moist sterile dressings until more definitive care can take place.
Since IV fluids are the primary therapy for hypovolemic shock, they will be considered as pharmacologic agents for this discussion.
Drug Treatments of First Choice
Dextrose-in-water solutions may be appropriate for uncomplicated dehydration caused by water deprivation, but isotonic crystalloid (sodium-containing) solutions should be used for forms of circulatory insufficiency that are associated with hemodynamic instability. In the latter situation, IV solutions with sodium concentrations approximating normal serum sodium values usually are indicated because they cause more expansion of the intravascular and interstitial spaces compared with dextrose solutions (Table 14-3). Lactated Ringer’s and normal saline solutions are examples of such crystalloid solutions that frequently need to be administered in large volumes when given to patients with more severe forms of hypovolemia. A “large” amount of fluid does not mean a single bolus volume typically used as fluid challenge in a critically ill patient. An isolated bolus (e.g., 250 to 500 mL) in a young adult trauma patient is unlikely to cause a substantial change in blood pressure or acid–base balance.8 Therefore, multiple fluid boluses usually are often needed in such patients to achieve hemodynamic stability in the perioperative period. On the other hand, overly aggressive fluid administration should be avoided, especially in patients with heart failure or impending pulmonary edema. In a randomized trial involving patients with acute lung injury and radiographic presence of pulmonary edema, a more conservative fluid management strategy led to significantly fewer ventilator-free days and days not spent in an ICU (P <0.001).9
TABLE 14-3 Fluid Distribution and Major Indicationsa
Published Guidelines or Treatment Protocols Recommendations for shock associated with trauma have been published as part of the Advanced Trauma Life Support (ATLS) course (http://www.facs.org/trauma/atls/).10 In the past, the ATLS guidelines were derived more from consensus of expert participants than evidence, but this has changed in more recent revisions. Guidelines for prehospital fluid administration in patients with trauma have been published by the Eastern Association for the Surgery of Trauma (EAST).11 Other evidence relative to fluid choice for resuscitation is available from systematic reviews,12–14 a guideline for perioperative fluid resuscitation,15 and a guideline pertaining to burn shock resuscitation.16 Taken as a whole, the recommendations from all of these sources are consistent in that isotonic (or near isotonic) crystalloid solutions are the initial fluid of choice for resuscitation in hypovolemic shock (Table 14-4).
TABLE 14-4 Summary of Evidence for Choice of Plasma Expander for Hypovolemic Shock
General Information Reporting Efficacy and Safety The choice between normal saline and lactated Ringer’s solutions for hypovolemia is largely based on clinician preference and adverse effect concerns (Table 14-5). Traditionally, lactated Ringer’s solution has been recommended for patients with hemorrhage because it is unlikely to cause the hyperchloremic metabolic acidosis that is seen with infusions of large volumes of normal saline. More recently, concerns have been raised relative to the proinflammatory effects (e.g., neutrophil activation) of the D-isomer form of lactate that is contained along with the L-isomer in commercially available racemic isomer solutions. There are advocates for the use of lactated Ringer’s solution containing only L-isomer lactate, particularly for more severe forms of hemorrhagic shock, since it avoids the proinflammatory effects of the racemic solution, while avoiding the hyperchloremia associated with normal saline.17 Additionally, other substitutes for racemic lactate such as ketone or pyruvate have shown beneficial effects on neutrophil activation and gene expression in vitro and are the subject of ongoing studies.
TABLE 14-5 Adverse Effects of Plasma Expanders: Crystalloids
Although lactated Ringer’s solution does contain lactate, it does not cause substantial elevations in circulating lactate concentrations when used as a resuscitation solution.18 Once adequate plasma volume has been restored by fluid administration, the body can readily clear the blood of the excess lactate that has accumulated from both anaerobic metabolism and lactated Ringer’s solution. However, blood samples for lactate determinations drawn through catheters (arterial and venous) that have not been cleared appropriately may have spurious increases or decreases in lactate concentrations because of retained lactated Ringer’s and nonlactated solutions (e.g., varying concentrations of dextrose-in-water or sodium chloride), respectively.19 Therefore, blood samples for lactate concentration determinations should be drawn from a catheter that has been cleared adequately (e.g., 5 mL) of infusate after temporarily stopping the fluid infusion.
Alternative Drug Treatments
A number of pharmacologic therapies show promise in animal models of shock, but few demonstrate success in subsequent trials involving patients with shock. In large part this is a result of the lack of acceptable animal models of shock that mimic the pathophysiology of patients. In cases in which a relevant animal model is available, care must be taken when extrapolating the information to forms of shock other than the one under study. This may be the problem with naloxone, which has been shown to raise blood pressure in some studies of shock but not in others.
While research continues on medications that improve oxygen transport, optimize oxygen utilization, and reduce reactive oxygen species and reperfusion injuries, fluids remain the mainstay of therapy for shock. Hypertonic sodium chloride solutions have been used as an alternative to isotonic crystalloid solutions for hypovolemic shock in some studies. By causing redistribution (i.e., pulling fluid) from the intracellular space, hypertonic solutions cause rapid expansion of the intravascular compartment, which is essential for vital organ perfusion. In head-injured patients, it has been postulated that this redistribution should decrease intracranial pressure because the vessels of the brain are more impermeable to sodium ions than are vessels in other areas of the body. Additionally, hypertonic sodium chloride solutions have beneficial immunomodulating actions when compared with more isotonic solutions in experiments with animals, although these actions have not always translated into similar beneficial effects in patients.
From a safety standpoint, hypertonic sodium chloride is considered to be a high-risk concentrated electrolyte solution. Potential dosing and administration errors and related adverse events can occur when hypertonic sodium solution is ordered and administered by clinicians relatively unfamiliar with its use. Potential adverse events include cellular crenation and damage caused by the dramatic fluid shifts associated with hypernatremia, hyperchloremic metabolic acidosis from hyperchloremia, and peripheral vein destruction from high osmolality. The osmolarity of 3% sodium chloride is 1,026 mOsm/L. Although there are some notable exceptions (e.g., peripheral parenteral nutrition solutions often approach 1,000 mOsm/L), IV solutions with osmolarity values above 600 mOsm/L are usually recommended for administration by central lines. In the limited number of studies conducted in humans to date, adverse effects related to hypertonic sodium solutions have been uncommon and apparently of little clinical importance. Larger-molecular-weight solutions (i.e., >30,000 Da) known as colloids have been recommended in conjunction with or as replacements for crystalloid solutions, although their use is controversial. The major theoretical advantage of these compounds is their prolonged intravascular retention time compared with crystalloid solutions. In contrast to isotonic crystalloid solutions that have substantial interstitial distribution within minutes of IV administration, colloids remain in the intravascular space for hours or days, depending on factors such as capillary permeability. Examples of colloids used as plasma expanders in the United States include albumin, hydroxyethyl starch, and dextran. Albumin is known as a monodisperse colloid because all its molecules are of the same molecular size and weight (∼67,000 Da), whereas hydroxyethyl starch and dextran solutions are polydisperse compounds with molecules of varying molecular size that are roughly proportional to molecular weight (weight-averagedmolecular weights of 600,000 Da [range 450,000 to 800,000 Da] for 6% hetastarch in normal saline 450/0.75, 670,000 Da [range 450,000 to 800,000 Da] for 6% hetastarch in lactated electrolyte 670/0.75, 130,000 Da [range 110,000 to 150,000 Da] for 6% tetrastarch in normal saline 130/0.4, 40,000 Da [range 10,000 to 90,000 Da] for dextran 40, or 70,000 to 75,000 Da [range 20,000 to 200,000 Da] for dextran 70 or dextran 75, respectively). In light of these differences, colloid comparisons are based on weight-averaged ([number of molecules at each weight × particle weight]/total weight of all molecules) or number-averaged (arithmetic mean of all particles’ weights) molecular weight.20 The size and weight differences of the colloids have important implications for the distribution of the products because lower-molecular-weight substances are retained in the intravascular space for a shorter period of time as a result of more rapid leakage across the vessel membrane. The theoretical benefit common to all colloids is based on their increased molecular weight (average molecular weight in the case of hydroxyethyl starch and dextran) that corresponds to increased intravascular retention time in the absence of increased capillary permeability compared with crystalloids. Even in patients with intact capillary permeability, the colloid molecules eventually will leak through the membrane. In the case of albumin with a distribution half-life of 15 hours in normal subjects, approximately 60% of administered albumin molecules (and associated fluid) would be shifted to the interstitial space within 3 to 5 days of exogenous administration. In patients with altered permeability (e.g., acute respiratory distress syndrome), the leakage of albumin from the intravascular to the interstitial space may occur within hours, not days. The primary adverse effect concern of all colloids is fluid overload, which is an extension of their pharmacologic action. Another adverse effect of increasing concern is renal dysfunction that seems to be related to hyperoncotic (e.g., 25%) albumin and other starch and dextran products. The mechanism of this adverse effect may be related to alteration of normal glomerular oncotic pressure differences or formation of lesions in the kidney.21
There is no widespread agreement on the upper limit of osmolarity for hypertonic sodium solutions that are given by peripheral vein infusion.
Albumin is available in 5% and 25% concentrations. Plasma protein fraction has oncotic actions similar to a 5% albumin solution, which is not surprising because albumin is the predominant protein in this product. When given in equipotent amounts, albumin is much more costly than crystalloid solutions. Additionally, the 5% and 25% albumin solutions typically are priced such that no cost saving is associated with dilution of the 25% product to make a 5% concentration. In general, dilution should be avoided because of the possibility of preparation errors; cases of hemolysis and death have occurred when 25% albumin was inappropriately diluted with sterile water for injection, causing a dramatic lowering of effective osmolarity. The 5% albumin solution is relatively iso-oncotic, which means that it does not pull fluid into the compartment in which it is contained. In contrast, 25% albumin is referred to as hyperoncotic albumin because it tends to pull fluid into the compartment containing the albumin molecules. In general, the 5% albumin solution is used for hypovolemic states. The 25% solution should not be used for acute circulatory insufficiency unless it is used in combination with other fluids or it is being used in patients with excess total body water but intravascular depletion as a means of pulling fluid into the intravascular space. An example of the latter condition is cirrhosis with ascites in which total body water is substantially increased, but the patient is hypotensive as a consequence of lack of intravascular volume. To justify this use of hyperoncotic albumin from a cost-effectiveness standpoint presumes that there is evidence of adverse effects associated with the excess water (e.g., interstitial fluid accumulation in the lungs) and that the albumin remains in the intravascular space long enough to be of benefit. Albumin has a variety of functions beyond plasma expansion, such as binding properties, inflammatory gene modification, and antioxidant and free radical scavenging effects, which have been used to justify its administration instead of less expensive crystalloid or other colloid products. Although appealing theoretically, improved patient outcomes related to these properties have not been documented in adequately powered, randomized controlled trials. Additionally, the clinician must realize that the properties of commercially available albumin products are not biologically identical to those of native albumin. For example, denaturation of the products may lead to inefficient binding and decreased oncotic activity.
Hydroxyethyl starch products have been developed as synthetic alternatives to albumin that is derived through the fractionation of donated human blood. The various products are differentiated by two numbers, one for the average mean molecular weight and one for the degree of hydroxyethyl substitution of glucose. For example, hetastarch is expressed as 450/0.7 based on weight and substitution, respectively. Most of the trials comparing albumin with hydroxyethyl starch products for volume expansion were inadequately powered and found no significant differences in clinically important outcomes (e.g., mortality). Two large randomized trials have directly compared hydroxyethyl starch products with crystalloid solutions for intravascular expansion. Although these trials used newer, low-molecular-weight (140), low-substitution (0.4 or 0.42) starch products, hemostasis and renal function problems noted in older trials involving high-molecular-weight, high-substitution starch products were found, suggesting these are class adverse effects. One of these large trials (Scandinavian Starch for Severe Sepsis/Septic Shock, also known as the 6S trial) found significantly higher rates of renal replacement therapy, red blood cell transfusions, and 90-day mortality in patients receiving hydroxyethyl starch versus a Ringer’s acetate solution.22 The second trial (referred to as the CHEST trial) involved 7000 general intensive care unit patients, making it the largest randomized study to date involving a starch product. As in the other large trial, patients in the hydroxyethyl starch group required significantly more renal replacement therapy versus patients receiving normal saline, but the 90-day mortality rates were similar.23 Possible explanations for the lack of a mortality difference include the relatively low overall mortality that might be related to the exclusion criteria (e.g., patients unlikely to survive), or to the use of normal saline as a control solution since saline has a high concentration of chloride ion and a low strong ion difference compared to plasma.
Hydroxyethyl starch may aggravate bleeding through mechanisms specific to this colloid (e.g., decreased factor VIII/von Willebrand factor). These mechanisms have not been well elucidated and often are difficult to distinguish from the dilutional effects on clotting factors caused by all plasma expanders; however, the risk of coagulopathy appears to be related to increasing doses and durations of administration.22 Renal dysfunction associated with hydroxyethyl starch products may also be a function of dose and duration of administration. Regardless of potential mechanisms, the FDA considers the serious adverse effects of the hydroxyethyl starch products to be class effects that warrant changes to product labeling. The changes include a boxed warning that states these products are contraindicated in critically ill patients. Additional warnings have also been added about excessive bleeding when used in patients undergoing cardiopulmonary bypass. Hydroxyethyl starch may cause elevations in serum amylase concentrations but does not cause pancreatitis.
The mechanisms by which hydroxyethyl starch products cause bleeding and acute kidney injury have yet to be fully elucidated.
Dextran 40, dextran 70, and dextran 75 are available for use as plasma expanders in the United States. The numbers refer to the average molecular weight of the solutions. In general, dextran solutions are not used as often as albumin for plasma expansion because of a lack of adequately powered randomized trials, and because of concerns related to aggravation of bleeding (i.e., anticoagulant actions related to inhibiting stasis of microcirculation), acute kidney injury, and anaphylaxis that is more likely to occur with the higher-molecular-weight solutions. There are few comparative trials involving the dextran solutions, but the intravascular expansion within hours after infusion is approximately equal to the amount of dextran infused. Apart from the acute kidney injury and bleeding associated with starch and dextran products, adverse effects associated with colloids generally are extensions of their pharmacologic activity (Table 14-6).
TABLE 14-6 Adverse Effects of Plasma Expanders: Colloids
From a historical perspective, the so-called crystalloid versus colloid debate was intensified when a meta-analysis by the well-respected Cochrane group found an overall increase in mortality associated with albumin using pooled results of randomized investigations.24 The meta-analysis involved 30 randomized trials with 1,419 patients (relative risk of death with albumin vs. no administration or crystalloid administration, 1.68; 95% confidence interval [CI], 1.26 to 2.23). For hypovolemia (caused by blood loss in the majority of studies), the risk of death associated with albumin administration was not quite statistically significant (relative risk, 1.46; 95% CI, 0.97 to 2.22). With the notable exception of trauma patients, a subsequent and more comprehensive systematic review did not find increased mortality attributable to albumin.25 Furthermore, a landmark investigation involving almost 7,000 critically ill patients (conducted after the previously mentioned meta-analyses) did not find statistically significant differences in 28-day mortality between patients resuscitated with either normal saline or 4% albumin.26 As in the previous meta-analysis, there was a trend toward increased mortality in patients with trauma, which became statistically significant (P = 0.003) when analyzed at 24 months in a subset of patients with traumatic brain injury.27 This multicenter, randomized, double-blind investigation, referred to as the Saline versus Albumin Fluid Evaluation (SAFE) study, involved a heterogeneous group of ICU patients and was not sufficiently powered to look at various subsets, so clinicians must be cautious when extrapolating the results to more specific patient populations. With this caution in mind, this trial provides strong evidence that crystalloid solutions should be considered first-line therapy in patients with hypovolemic shock.
The colloids are expensive solutions. Therefore, it is difficult to justify the additional cost of colloidal products unless the benefit-to-risk ratio is substantially greater than that associated with inexpensive crystalloid solutions. This does not appear to be the case based on randomized controlled studies and meta-analyses comparing colloid and crystalloid solutions for acute circulatory insufficiency. While the use of albumin in specific patient populations (e.g., septic shock) is still debated, the documented adverse effect profile of hydroxyethyl starch products and the lack of adequately powered trials for dextran products renders them all unsuitable for use in critically ill patients including those with shock.
In contrast to other forms of shock such as anaphylactic or septic, medications are a distant alternative to the primary therapy for hypovolemic shock, fluids. In hypovolemic shock, peripheral resistance is high due to compensatory mechanisms aimed at maintaining tissue perfusion. Early or overzealous use of vasopressors in lieu of fluids may exacerbate this resistance to the point that flow is stopped. Therefore, vasoactive agents that dilate the peripheral vasculature such as dobutamine are preferred if the blood pressure is stable and high enough to tolerate the vasodilation. Vasopressors are only used as a temporizing measure or as a last resort when all other measures to maintain perfusion have been exhausted.28 Because vasopressors have such a limited role in hypovolemic shock, there are very few studies that compare various agents. In one of the few studies that included patients with hypovolemic shock, norepinephrine and dopamine had similar effects on mortality, but dopamine was associated with more adverse effects, particularly atrial fibrillation.29
The need for immediate treatment of hemorrhagic circulatory insufficiency with plasma expanders (i.e., crystalloids or colloids) seems obvious, but no large, well-controlled trials conducted in humans have supported this practice. To the contrary, evidence suggests that fluid resuscitation beyond minimal levels (i.e., mean arterial pressure >60 mm Hg) is harmful in patients with penetrating abdominal trauma due to hemodilution and clot destabilization. One prospective study involving 598 adult patients with gunshot or stab wound injuries to the torso and systolic blood pressure measurements of 90 mm Hg or less found that delayed fluid resuscitation until operation was associated with increased survival and discharge from the hospital (P = 0.04).30 Since concerns were expressed about the comparability of the immediate and delayed resuscitation groups, particularly because true randomization did not take place, a follow-up randomized trial was conducted to verify the findings. There were no differences in survival (four deaths in each group) in the second trial regardless of whether systolic blood pressure was titrated to >100 mm Hg or to 70 mm Hg.31 Both studies were conducted in populated urban areas with approximately 2 hours from the time of injury to operation. Therefore, the results may not be applicable to rural areas with extended transport times. There also is a concern in applying the results of these investigations to patients with certain kinds of single-system injuries, particularly head trauma, where cerebral perfusion pressure is of primary importance. Although the applicability of these studies to other populations and settings is debatable, the presumption of benefits from immediate plasma expansion in all preoperative patients with circulatory insufficiency caused by hemorrhage is no longer valid. Instead, the initial priority should be surgical control of the bleeding source; until this is possible, fluids should be given in small aliquots to yield a palpable pulse and to maintain mean arterial pressures no more than 60 mm Hg and systolic pressures no more than 90 mm Hg based on accurate measurements (e.g., arterial monitoring).
Beneficial outcome data attributable to hypertonic solutions are lacking. Most of these studies were conducted in prehospital and emergency department settings using 250 mL of 7.5% sodium chloride with or without 6% dextran 70. For example, a double-blind, randomized controlled trial involving 229 patients with hypotension and severe brain injury demonstrated no significant differences in neurologic function at 6 months when 250 mL of 7.5% saline or lactated Ringer’s solution was administered as part of a prehospital resuscitation regimen.32 Part of the explanation for this finding may be related to supplemental crystalloid fluids that were given routinely to patients in both the treatment and control groups, which probably would increase the number of patients needed to demonstrate a statistically significant difference in mortality.
In order to address ongoing questions of efficacy, the National Heart, Lung, and Blood Institute evaluated hypertonic sodium chloride solutions with or without a colloid (i.e., 7.5% sodium chloride or 7.5% sodium chloride in 6% dextran 70) for prehospitalized trauma patients with shock and severe traumatic brain injury in two trials conducted by a network of sites known as the Resuscitation Outcomes Consortium (ROC). Both the parallel trials were stopped when it was determined that the hypertonic sodium chloride solutions were no better than normal saline and further enrollment would not change the 33 outcomes.33,34 Therefore, normal saline is the fluid of choice since it is equal in efficacy with a lower risk of adverse effects compared with hypertonic solutions that are high-risk electrolyte solutions. Given their relatively poor intravascular expansion and association with poor outcome in animal models of closed head injury, hypotonic solutions should be avoided in this population.
In addition to crystalloid solutions, colloids have been used for plasma expansion in trauma patients with perioperative circulatory insufficiency. No large randomized studies have compared crystalloids and colloids for circulatory insufficiency in trauma patients. Until such studies are performed, there is no compelling reason to suspect that colloids have any substantial clinical benefits beyond crystalloids in these patients given the results of previous trials and systematic reviews performed in more general critical care populations. Further, bleeding and renal injury concerns for both starch and dextran products precludes their use in critically ill trauma patients.
The preceding discussion dealt primarily with acute circulatory insufficiency, but there are other considerations with regard to fluid replacement in other patients undergoing surgical procedures. Preoperative fluid deficits in patients undergoing minor procedures may be associated with increased perioperative morbidity, some of which (e.g., drowsiness, dizziness) may be reduced by appropriate fluid administration prior to surgery. However, care must be taken to avoid overhydration in the perioperative period because excess fluid will lead to weight gain and decreased pulmonary function. Some evidence suggests that fluid restriction on the day of surgery may reduce postoperative morbidity in patients undergoing major surgical procedures. In one randomized, multicenter trial, use of a restricted intraoperative and postoperative IV fluid protocol led to significantly fewer cardiopulmonary (7% vs. 24%; P = 0.007) and wound (16% vs. 31%; P = 0.04) complications.35 As the preceding discussion indicates, the benefits and risks of fluid administration in the perioperative period are not just a function of too little or too much fluid but involve other patient- and procedure-related issues.
Another consideration in the patient with penetrating injuries or surgery is the potential need for blood product administration (Table 14-7) to replace oxygen-carrying and clotting functions. Although a small group of trauma patients respond to the initial fluid bolus and remain stable, most patients respond initially and then deteriorate. The latter patients, as well as patients undergoing blood loss associated with surgery, frequently need blood components such as packed red blood cells. In the case of the latter component, red blood cells contain hemoglobin that delivers oxygen to tissues. Neither crystalloids nor colloids perform this function.
TABLE 14-7 General Indications for Blood Products in Acute Circulatory Insufficiency Due to Hemorrhagea
Administration of excessive blood products may be counterproductive. In the case of red blood cells, attempts to raise the hematocrit to high–normal or supranormal concentrations may decrease oxygen delivery by increasing blood viscosity. Additionally, there are immunomodulatory concerns with red blood cell administration. Although there is no optimal hematocrit value for all patients, a minimum hematocrit of 30% (0.30) (equivalent to a hemoglobin concentration of 10 g/dL [100 g/L; 6.21 mmol/L]) traditionally has been used as the threshold for transfusion, particularly in patients at risk for ischemia, such as those with CAD. Use of a more liberal transfusion strategy has been curtailed in many institutions with the publication of a randomized, multicenter trial involving critically ill patients that found 30-day mortality to be similar whether patients were transfused at a hemoglobin concentration less than 7 or 10 g/dL (70 to 100 g/L; 4.34 to 6.21 mmol/L) (18.7% vs. 23.3%, respectively; P = 0.11).36 The mortality during hospitalization was significantly lower in the restrictive group (22.2% vs. 28.1%; P = 0.05). Although the investigators were cautious about extrapolating the results of this investigation to patients with myocardial ischemia, a subsequent study performed in patients undergoing cardiac surgery found similar results.37 With the exception of the critically ill or perioperative patient with acute exsanguination, there is little justification for a liberal transfusion strategy based solely on hemoglobin concentrations.
Blood products have risks beyond immunomodulation. There is the rare but important risk of virus transmission (e.g., human immunodeficiency virus [HIV], hepatitis). Citrate that is added to stored blood to prevent coagulation may bind to calcium, resulting in hypocalcemia, although potassium and phosphate concentrations often are elevated in stored blood, particularly when hemolysis has occurred during storage. In patients receiving large amounts of blood, prophylactic calcium administration may be warranted until levels are available. Other issues that must be considered with blood product administration include monitoring for transfusion-related reactions and attention to appropriate warming, particularly when large volumes are given to pediatric patients, because hypothermia is associated with increased fluid requirements and mortality.
Since its commercial release in the United States, recombinant factor VIIa has been used for a variety of off-label uses related to trauma and bleeding. For example, in patients with massive blood loss a cocktail of cryoprecipitate, platelets, and recombinant factor VIIa has been suggested to rapidly attain hemostasis. These more severe forms of blood loss are a function of not only the type of injury but also factors such as medications (e.g., aspirin, Coumadin, clopidogrel, enoxaparin) and disease states that impair normal coagulation. Large well-controlled trials are needed to define the role of recombinant factor VIIa in clinical practice given its high cost and potential thromboembolic complications. Concerns with its use in trauma patients are issues related to appropriate dose, timing, and diminished effectiveness in patients with acidosis and severe hypothermia. Evidence of efficacy in a general trauma population that would offset these concerns is lacking. In the largest randomized controlled trial conducted to date that enrolled patients with penetrating and blunt trauma, factor VIIa did not decrease mortality compared with placebo when the trial was prematurely terminated due to futility.38
The periodic shortages, high costs, and adverse effect concerns related to blood products have prompted investigations of alternative “bloodless” strategies. In addition to the use of more restrictive transfusion thresholds, as mentioned previously, these strategies have included hemoglobin-based oxygen carriers and perfluorocarbon compounds to deliver oxygen to tissues. Other strategies have aimed at reducing blood loss through the use of improved procedural and surgical techniques, as well as the administration of hemostatic medications. The only hemostatic medication with a proven mortality benefit is the antifibrinolytic agent, tranexamic acid. The best evidence for efficacy was data from a multicenter trial involving more than 20,000 adult trauma patients with significant bleeding (or risk for significant bleeding) who were randomized to IV tranexamic acid (1 g over 10 minutes followed by 1 g over 8 hours by infusion) or matching placebo within 8 hours of injury.39 There was a significant reduction in all-cause mortality with tranexamic acid compared with placebo (14.5% vs. 16%, P = 0.0035) with no increase in vascular or other adverse events. While additional data are still needed in specific subpopulations such as patients with traumatic brain injuries, this study is relatively unique in that an intervention apart from surgery and blood product administration was demonstrated to reduce mortality.
Patients with Thermal Injuries
There are a number of formulas for estimating fluid requirements in thermally injured patients, but there is little reason to choose one over another based on well-controlled studies. In general, the amount of loss corresponds to the size of the thermal injury. Guidelines recommend approximately 2 to 4 mL/kg of isotonic fluid (lactated Ringer’s solution) for each percent burn can be used for calculating the expected fluid requirements for the first 24 hours after the burn.16 For example, a 60-kg person with 30% body surface area (BSA) burns is expected to require 5,400 to 7,200 mL of fluid over the initial 24 hours. Regardless of the calculated deficit, fluids should be administered until adequate tissue perfusion has been documented (e.g., maintenance of urine output of 0.5 to 1 mL/kg in adults) or adverse effects (e.g., pulmonary edema) occur. Crystalloids are preferred as initial therapy for burn victims because there is no substantial evidence that colloids mobilize edematous fluid, and there is a theoretical concern that extravascular fluid accumulation might be prolonged by the oncotic actions of albumin and other colloid products that have leaked through vessel walls. Additionally, there is no evidence that colloids reduce mortality in patients with thermal injuries and there is a concern that hydroxyethyl starch and dextran products might even increase mortality through deleterious effects on coagulation and renal function. Some novel therapies for thermal resuscitation have been studied, although larger confirmatory trials are needed prior to use apart from research protocols. For example, in a prospective study involving patients with >30% BSA burns, antioxidant therapy with extremely high doses of IV vitamin C (66 mg/kg/h for 24 hours) reduced resuscitation fluid requirements and wound edema.40 The proposed mechanism is reduction in free radical–induced increases in capillary permeability.
At this time there is little genetic/genomic information that is available to guide personalized pharmacotherapy in patients with hypovolemic shock. Further, as stressed throughout this chapter, fluids are by far the first choice of therapy in conjunction with other definitive interventions such as surgery for traumatic injuries. Nevertheless, there are individual factors that may influence the specific fluid being administered. For example, the lower chloride concentration in lactated Ringer’s would usually make it preferred over normal saline in patients with a hyperchloremic metabolic acidosis, while the increased osmolarity of normal saline would usually make it preferred over lactated Ringer’s in a patient with increased intracranial pressure.
Some clinicians believe that hypertonic solutions should be used to lower intracranial pressure in patients with head injuries.
The appropriate use of invasive hemodynamic monitoring tools, such as right-sided heart catheterization in patients with hypovolemic shock, is controversial.
EVALUATION OF THERAPEUTIC OUTCOMES
Monitoring of the Pharmaceutical Care Plan
One form of monitoring that may take place in the emergency and operating rooms, as well as in the ICU, requires placement of a central venous pressure (CVP) line. Monitoring of CVP provides the clinician with a somewhat insensitive yet useful estimate of the relationship between increased right atrial pressure and cardiac output. A protocol that used a particular type of central catheter to perform continuous monitoring of central venous oxygen saturation in conjunction with so-called goal-directed therapy in the first 6 hours of patient arrival in an urban emergency department resulted in decreased mortality compared with standard monitoring (30.5% vs. 46.5%; P = 0.009).41 However, the patients in this study had severe sepsis and septic shock, so the results might not be applicable to other forms of shock with different pathophysiologic considerations. For example, in hemorrhagic shock due to trauma, the most important intervention is surgical control of bleeding, and anything that delays this control is likely to increase, not decrease, mortality. Until additional studies have been performed, it would be premature to mandate goal-directed therapy with the associated central venous monitoring in patients with nonseptic forms of shock, particularly shock due to blood loss. In fact, so-called “upstream” measurements of perfusion such as CVP are not a useful guide for fluid management in hospitalized patients, and are being replaced by “downstream” markers such as urine output and lactate levels that are more likely to reflect end-organ dysfunction.42 A more complete discussion of invasive and noninvasive hemodynamic monitoring is given in Chapter 1.
The most appropriate, cost-effective, and practical parameter(s) for monitoring adequacy of fluid resuscitation in shock is unresolved.
A number of laboratory tests are indicated for subacute monitoring of shock in the ICU setting. These include a renal battery for assessing possible electrolyte alterations and kidney perfusion (e.g., BUN and creatinine). Among other things, a complete blood count will enable assessment of possible infection (white blood cell count), oxygen-carrying capacity of the blood (hemoglobin, hematocrit), and ongoing bleeding (hemoglobin, hematocrit, and platelet count). The PT or international normalized ratio and partial thromboplastin time (PTT) will give an indication of the ability of the blood to clot because, in the case of hemorrhagic shock, clotting factors are lost and diluted. An increasing lactate concentration (arterial, mixed venous, or central venous), an increasing arterial base deficit, or a decreasing bicarbonate concentration are global markers indicative of inadequate perfusion leading to anaerobic metabolism with accumulation of lactic acid. Although the value of these surrogate markers for improving patient outcomes is more controversial, they are considered traditional end points of resuscitation in certain populations such as trauma patients. Other tests may be indicated if organ dysfunction is likely. For example, when blood flow to the liver is interrupted because of sustained hypotension, a condition known as shock liver may occur. In this condition, the levels of transaminases on a liver panel may be markedly elevated in the first couple of days after marked hypotension, although the concentrations should decrease over time. Along with laboratory testing, a more extensive history can be obtained during the subacute monitoring period.
The value of pulmonary artery catheters (also known as right-sided heart or Swan-Ganz catheters) has been debated hotly since their introduction. Such catheters are placed to obtain various oxygen-transport variables, some of which cannot be determined reliably from peripheral or other central vessels. The debate was intensified when early studies suggested improved outcomes when cardiac output and other oxygen-transport variables were raised to supranormal levels, the monitoring of which required placement of a pulmonary artery catheter. The controversy led to consensus conferences and workshops, the development of organizational guidelines, and the publication of a meta-analysis (which found a statistically significant reduction in morbidity using pulmonary artery catheters to guide therapy).43 Ultimately, a large randomized controlled trial involving pulmonary artery catheters was conducted in high-risk surgical patients.44 The trial involved 1,994 patients. The mortality was almost identical for the catheter and control groups (7.8% vs. 7.7%; 95% CI, 2.3 to 2.5). There were no episodes of pulmonary embolism in the catheter group and eight episodes in the control group (P = 0.004). This trial is important not only because of the implications for high-risk surgical patients but also because it allows for the conduct of future trials in other patient populations without some of the ethical issues raised about such trials in the past.
Part of the concern regarding pulmonary artery catheterization relates to interpretation of its results by inexperienced practitioners. Studies in Europe and the United States found that one of two physicians incorrectly interpreted a tracing from a pulmonary artery catheter.45 This could explain some of the results of studies finding no benefits to pulmonary artery catheterization or, in some cases, worse outcomes in the pulmonary artery catheterization group by actions taken as a result of inaccurate measurements or misinterpretation of information obtained from the monitoring process.
Complications related to pulmonary artery catheter insertion, maintenance, and removal include damage to vessels and organs during insertion, arrhythmias, infections, and thromboembolic damage. To avoid the complications associated with pulmonary artery catheterization, other less invasive tools were developed to obtain similar information. For example, cardiac output determinations have been made by Doppler, bioimpedance, dye, and ionic dilution techniques, although such measurements would not provide other data that are obtained routinely with pulmonary artery catheters (e.g., left-sided heart filling pressure). Additionally, advances in pulmonary artery catheter technology that expand the information obtained from such monitoring (e.g., mixed venous oxyhemoglobin) are under investigation. However, given the lack of well-defined outcome data associated with pulmonary artery catheterization, its use is best reserved for complicated cases of shock not responding to conventional fluid and medication therapies.
Commonly measured and calculated hemodynamic and oxygen-transport indices associated with invasive monitoring are primarily global indicators of tissue perfusion. Attempts have been made to find regional and local indicators of hypoperfusion so that circulatory insufficiency could be treated before overt shock occurs. One focus of recent research has been monitoring modalities involving the GI tract. Although the literature is fairly consistent concerning low gastric intramucosal pH (pHi) values being predictive of death, pHi-guided therapy to decrease mortality has not been demonstrated.46 Additionally, a number of technical considerations remain to be resolved when using pHi or, more recently, capnometry (luminal PCO2 tonometry) for monitoring and therapy. Despite these concerns, measures of regional tissue oxygenation continue to be investigated through a variety of novel monitoring techniques.
In addition to regional monitoring of tissue perfusion, local methods of monitoring are being studied. For example, subcutaneous measurement of tissue oxygen pressure shows promise in preliminary investigations. Regional and local measurements likely will not replace more global indicators of perfusion; rather, the methods will complement each other.
Proper attention to monitoring of plasma volume must be continued into the intraoperative and postoperative periods. A number of neurohormonal changes take place that affect urine output, and patients may have substantial third spacing of fluid depending on the operation and preexisting conditions. Furthermore, postoperative patients are prone to hyponatremia from renal generation of electrolyte-free water and from antidiuretic hormone release. As in acute resuscitation, the administration of hypotonic solutions in the perioperative period does not prevent the decrease in extracellular volume that often occurs. Therefore, although excess fluid administration is to be avoided in the perioperative setting, isotonic crystalloid solutions should be used when fluids are indicated to prevent intravascular depletion and circulatory insufficiency. There is general agreement that the choice of crystalloid solution in the perioperative period should be either normal saline or a lactated Ringer’s (or equivalent) solution. However, there is substantial debate as to which of these two solutions is preferable since comparative studies have involved small numbers of patients.
Of the randomized studies comparing albumin with crystalloid solutions in the perioperative period, the majority found no statistically significant differences between groups. Any significant differences found involved isolated hemodynamic or respiratory variables with no obvious clinical correlates (e.g., duration of mechanical ventilation). Therefore, albumin cannot be recommended for the prevention or initial treatment of circulatory insufficiency, although its use may be appropriate in patients who are not responding to crystalloids and are developing problems such as interstitial fluid accumulation.
In contrast to many other forms of shock, vasoactive medications are not indicated in the initial therapy of hypovolemic shock but rather play an adjunctive role in patients who continue to have circulatory insufficiency after fluids have been maximized and volume-overload concerns exist. In a multicenter cohort study of blunt-injured patients with hemorrhagic shock, the use of vasopressors within 12 hours of injury was associated with significantly higher mortality at 24 hours (P = 0.001).47 With hypovolemia, the body’s natural response is to increase cardiac output and to constrict blood vessels to maintain blood pressure. There is no evidence that vasoactive medications improve outcome in patients with hypovolemic shock assuming that fluid therapy is adequate. However, once the cause of acute circulatory insufficiency has been stopped or treated and fluids have been optimized, some patients continue to have signs and symptoms of inadequate tissue perfusion. This may be caused by reperfusion injury. Although the search for a cryptogenic source (e.g., intraabdominal bleeding in a trauma patient) should continue, the clinician may need to administer vasoactive medications to improve perfusion.
Pressor agents such as norepinephrine and high-dose dopamine are to be avoided, if possible, because they may increase blood pressure at the expense of peripheral tissue ischemia. Some sources use stronger language and state that vasopressors are contraindicated in certain forms of shock (e.g., hemorrhagic). This does not help the clinician who is treating a patient with unstable blood pressure despite massive fluid replacement and increasing interstitial fluid accumulation. In such situations, inotropic agents such as dobutamine are preferred if blood pressure is adequate (e.g., systolic blood pressure ≥80 to 90 mm Hg) because they should not aggravate the existing vasoconstriction. The inotropic agents are justified by presumed inadequate cardiac output for the specific situation, although the measured values may be in the normal range.
When pressure cannot be maintained with inotropic agents or when inotropic agents with vasodilatory properties cannot be used because of inadequate blood pressure concerns, pressors may be required as a last resort. In general, the need for pressors is predictive of the development of MODS and increased length of hospital stay. Although the response to pressor agents may be variable in hypovolemic shock, there does not appear to be resistance as a consequence of altered receptor response, as is sometimes seen in patients with septic shock. Potent vasoconstrictors such as norepinephrine and phenylephrine should be given through central veins because of the possibility of extravasation and necrosis with peripheral administration.
In managing patients with hypovolemic shock, the clinician must be aware of potential adverse effects of medications being used for supportive care purposes. For example, some patients are particularly susceptible to the histamine release associated with morphine and may have substantial decreases in blood pressure. Sodium bicarbonate would seem to be a logical therapy in patients with shock who typically have a metabolic acidosis, but bicarbonate administration has not been shown to improve surrogate hemodynamic markers or patient outcomes and has known disadvantages such as the associated increase in arterial carbon dioxide levels and decrease in serum ionized calcium levels.48 Propofol is commonly used for sedation in the ICU, but it may cause substantial decreases in blood pressure. The initial dose of propofol should be substantially reduced or avoided in patients with hemorrhagic shock who may not be fully resuscitated.
A number of interesting treatments for shock are under investigation, including autotransfusion for removing harmful cytokines from the body. Various alternatives to conventional blood components also are being studied, such as stroma-free hemoglobin and perfluorocarbon compounds, as virus-free alternatives to red blood cell transfusion. Hopefully, these methods will be useful adjuncts to adequate volume replacement, which is the primary therapeutic intervention in managing acute circulatory insufficiency as a result of volume depletion.
CLINICAL BOTTOM LINE
Figure 14-4 is an algorithm that summarizes many of the treatment principles discussed in this chapter. The algorithm is an example of one approach to the adult patient presenting with hypovolemic shock. It presumes that initial rehydration attempts (i.e., outpatient or prehospital) were unsuccessful in restoring circulation. Obviously, modifications may be needed for patient-specific forms of hypovolemic shock. For example, in patients with severe traumatic brain injury albumin would be contraindicated as a plasma expander, while hypertonic sodium solution might be considered for its ability to lower elevated intracranial pressure without causing the diuresis associated with mannitol administration. Other limitations of the algorithm should be recognized, particularly the decisions to add or to substitute medication therapies when crystalloid solutions are not yielding desired results and when to perform pulmonary artery catheterization for more invasive monitoring. Medications become more important for the ongoing management of hypovolemic shock, but only when the patient is unresponsive to fluids (Fig. 14-5). Medications for more complicated cases of hemorrhagic shock should not detract from the primary effective resuscitative measure—surgical stabilization of bleeding.
FIGURE 14-4 Hypovolemia protocol for adults. Normal saline may be used instead of lactated Ringer’s solution. This protocol is not intended to replace or delay therapies such as surgical intervention or blood products for restoring oxygen-carrying capacity or hemostasis. For the resuscitation of patients with trauma prior to bleeding control, usually no more than 1 liter of crystalloid should be given initially in an attempt to use the minimal amount of fluid necessary to maintain perfusion and not exacerbate bleeding. If available, some measurements can be used in addition to those listed in the algorithm, such as mean arterial pressure or pulmonary artery catheter recordings. The latter can be used to assist in medication choices (e.g., agents with primary pressor effects may be desirable in patients with normal cardiac outputs, whereas dopamine or dobutamine may be indicated in patients with suboptimal cardiac outputs). Lower maximal doses of the medications in this algorithm should be considered when pulmonary artery catheterization is not available. See text for an in-depth discussion of these and other issues involved in this protocol. (CHF, congestive heart failure; LR, lactated Ringer’s solution.)
FIGURE 14-5 Ongoing management of inadequate tissue perfusion. Normal saline may be substituted for lactated Ringer’s solution in this figure. (LR, lactated Ringer’s solution.)
The algorithm in Figure 14-4 attempts to incorporate economic considerations. The institutional cost of 1 L of most crystalloid solutions is less than $1. Assuming that such fluids are used, the associated costs of personnel and equipment then become the primary economic considerations in the resuscitation of patients with hypovolemic shock. However, as mentioned, many clinicians recommend that colloid plasma expanders (e.g., albumin, hydroxyethyl starch, or dextrans) be used to replace some or all of the standard crystalloid solutions. Although the costs of these solutions vary, depending on contractual arrangements, in general, albumin solutions are more expensive than older hydroxyethyl starch and dextran products. All these solutions are markedly more costly than crystalloid solutions; in some cases, the differences are 50- to 100-fold, even when used in equipotent amounts. It is important to note that these cost minimization statements assume no differences in efficacy or toxicity between colloids and crystalloids when given in equipotent amounts. This is almost certainly not the case with respect to adverse effects of hydroxyethyl starch and dextran products. A cost-effectiveness analysis that takes into account adverse effects would be needed for the latter products and such an analysis would likely demonstrate they are not cost-effective versus crystalloids even if equipotent efficacy is presumed.
Because medications are not simply alternatives to crystalloids but rather are used when crystalloid therapy has been optimized, there is little reason to compare medication and fluid therapies from an economic perspective. Furthermore, there are no economic comparisons of the various inotropic and vasopressor medications used in the treatment of hypovolemic shock.
1. Murphy SL, Xu J, Kochanek KD. Deaths: Preliminary Data for 2010. National Vital Statistics Reports, Vol. 60, No. 4. Hyattsville, MD: National Center for Health Statistics, 2012.
2. Anonymous. QuickStats: Number of Heat-Related Deaths,* by Sex—National Vital Statistics System, United States, 1999–2010. MMWR Morb Mortal Wkly Rep 2012;61(36):729.
3. Duggan C, Fontaine O, Pierce NF, et al. Scientific rationale for a change in the composition of oral rehydration solution. JAMA 2004;291:2628–2631.
4. Vercueil A, Grocott MPW, Mythen MG. Physiology, pharmacology, and rationale for colloid administration for the maintenance of effective hemodynamic stability in critically ill patients. Trans Med Rev 2005;19:93–109.
5. Porter SC, Fleisher GR, Kohane IS, Mandl KD. The value of parental report for diagnosis and management of dehydration in the emergency department. Ann Emerg Med 2003;41:196–205.
6. Spandorfer PR, Alessandrini EA, Joffe MD, et al. Oral versus intravenous rehydration of moderately dehydrated children: A randomized, controlled trial. Pediatrics 2005;115:295–301.
7. Atherly-John YC, Cunningham SJ, Crain EF. A randomized trial of oral vs intravenous rehydration in a pediatric emergency department. Arch Pediatr Adolesc Med 2002;156:1240–1243.
8. Axler OA, Tousiganant C, Thompson CR, et al. Small hemodynamic effect of typical rapid volume infusions in critically ill patients. Crit Care Med 1997;25:965–970.
9. The National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006;354:2564–2575.
10. The ATLS Subcommittee, American College of Surgeons’ Committee on Trauma, and the International ATLS working group. Advanced trauma life support (ATLS®): The ninth edition. J Trauma Acute Care Surg 2013;74:1363–1366.
11. Cotton BA, Jerome R, Collier BR, et al. Guidelines for prehospital fluid resuscitation in the injured patient. J Trauma 2009;67:389–402.
12. Perel P, Roberts I, Ker K. Colloids versus crystalloids for fluid resuscitation in critically ill patients. Cochrane Database Syst Rev 2013, Issue 2. Art. No.: CD000567. DOI: 10.1002/14651858.CD000567.pub6.
13. Bunn F, Trivedi D. Colloid solutions for fluid resuscitation. Cochrane Database Syst Rev 2012, Issue 7. Art. No.: CD001319. DOI: 10.1002/14651858.CD001319.pub5.
14. Roberts I, Blackhall K, Alderson P, Bunn F, Schierhout G. Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane Database Syst Rev 2011, Issue 11. Art. No.: CD001208. DOI: 10.1002/14651858.CD001208.pub4.
15. Soni N. British consensus guidelines on intravenous fluid therapy for adult surgical patients—Cassandra’s view. Anesthesia 2009;64:235–238.
16. Pham TN, Cancio LC, Gibran NS. American Burn Association practice guidelines burn shock resuscitation. J Burn Care Res 2008;29:257–266.
17. Alam HB, Rhee P. New developments in fluid resuscitation. Surg Clin North Am 2007;87:55–72.
18. Didwania A, Miller J, Kassel D, et al. Effect of intravenous lactated Ringer’s solution infusion on the circulating lactate concentration: Results of a prospective, randomized, double-blind, placebo-controlled trial. Crit Care Med 1997;25:1851–1854.
19. Jackson EV, Wiese J, Sigal B, et al. Effects of crystalloid solutions on circulating lactate concentrations: 1. Implications for the proper handling of blood specimens obtained in critically ill patients. Crit Care Med 1996;24:1840–1846.
20. Grocott MPW, Mythen MG, Gan TJ. Perioperative fluid management and clinical outcomes in adults. Anesth Analg 2005;100:1093–1106.
21. Schortgen F, Giron E, Deye N, Brochard L. The risk associated with hyperoncotic colloids in patients with shock. Intensive Care Med 2008;34:2157–2168.
22. Perner A, Haase N, Guttormsen AB, et al. Hydroxyethyl starch 130/0.42 versus Ringer’s acetate in severe sepsis. N Engl J Med 2012;367:124–134.
23. Myburgh JA, Finfer S, Bellomo R, et al. Hydroxyethyl starch or saline for fluid resuscitation in intensive care. N Engl J Med 2012;367:1901–1911.
24. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: Systematic review of randomized controlled trials. BMJ 1998;317:235–240.
25. Choi PTL, Yip G, Quinonez LG, Cook DJ. Crystalloids vs. colloids in fluid resuscitation: A systematic review. Crit Care Med 1999;27:200–210.
26. The SAFE Study Investigators. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004;350:2247–2256.
27. The SAFE Study Investigators. Saline or albumin for fluid resuscitation in patients with traumatic brain injury. N Engl J Med 2007;357:874–884.
28. Hollenberg SM. Vasoactive drugs in circulatory shock. Am J Respir Crit Care Med 2011;183:847–855.
29. De Backer D, Biston P, Devriendt J, et al. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med 2010;362:779–789.
30. Bickell WH, Wall MJ, Pepe PE, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med 1994;331:1105–1109.
31. Dutton RP, Mackenzie CF, Scalea TM. Hypotensive resuscitation during active hemorrhage: Impact on in-hospital mortality. J Trauma 2002;52:1141–1146.
32. Cooper DJ, Myles PS, McDermott FT, et al. Prehospital hypertonic saline resuscitation of patients with hypotension and severe traumatic brain injury. JAMA 2004;291:1350–1357.
33. Bulger EM, May S, Brasel KJ, et al. Out-of-hospital hypertonic resuscitation following severe traumatic brain injury: a randomized controlled trial. JAMA 2010;304:1455–1464.
34. Bulger EM, May S, Kirby JD, et al. Out-of-hospital hypertonic resuscitation after traumatic hypovolemic shock: A randomized, placebo controlled trial. Ann Surg 2011;253:431–421.
35. Brandstrup B, Tonnesen H, Beier-Holgersen R, et al. Effects of intravenous fluid restriction on postoperative complications: Comparison of two perioperative fluid regimens. Ann Surg 2003;238:641–648.
36. Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409–417.
37. Hajjar LA, Vincent JL, Galas FRBG, et al. Transfusion requirements after cardiac surgery: The TRACS randomized controlled trial. JAMA 2010;304:1559–1567.
38. Hauser CJ, Boffard K, Dutton R, et al. Results of the CONTROL trial: Efficacy and safety of recombinant activated factor VII in the management of refractory traumatic hemorrhage. J Trauma 2010;69:489–500.
39. CRASH-2 Trial Collaborators. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant hemorrhage (CRASH-2): A randomised, placebo-controlled trial. Lancet 2010;376:23–32.
40. Tanaka J, Matsuda T, Miyagantani Y, et al. Reduction of resuscitation fluid volumes in severely burned patients using ascorbic acid administration. Arch Surg 2000;135:326–331.
41. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic-shock. N Engl J Med 2001;345:1368–1377.
42. Marik PE, Baram M, Vahid B. Does central venous pressure predict fluid responsiveness? Chest 2008;134:172–178.
43. Ivanov R, Allen J, Calvin JE. The incidence of major morbidity in critically ill patients managed with pulmonary artery catheters: A meta-analysis. Crit Care Med 2000;28:615–619.
44. Sandham JD, Hull RD, Brant RF, et al. A randomized, controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003;348:5–14.
45. Ginosar Y, Thijs LG, Sprung CL. Raising the standard of hemodynamic monitoring: Targeting the practice or the practitioner? Crit Care Med 1997;25:209–211.
46. Gomersall CD, Joynt GM, Freebairn RC, et al. Resuscitation of critically ill patients based on the results of gastric tonometry: A prospective, randomized, controlled trial. Crit Care Med 2000;28:607–614.
47. Sperry JL, Minei JP, Frankel HL, et al. Early use of vasopressors after injury: Caution before constriction. J Trauma 2008;64:9–14.
48. Boyd JH, Walley KR. Is there a role for sodium bicarbonate in treating lactic acidosis from shock? Curr Opin Crit Care 2008;14:379–383.