Chapter 47. Fluid Management

Relative and absolute hypovolemia complicate many clinical conditions, and fluid therapy is a cornerstone of acute critical illness management. The clinician is constantly tasked with assessment of volume status, the need for fluid therapy, and selection of the appropriate fluid and dose guided to a suitable endpoint. Timely fluid therapy maintains macrocirculatory and microcirculatory support and reduces morbidity and mortality.1–3 In contrast, both under-resuscitation and over-resuscitation adversely affect outcome; inadequate resuscitation risks leaving a patient in compensated shock, and overly aggressive fluid administration results in volume overload without improving oxygen delivery and is associated with worse clinical outcomes.4,5 A thorough understanding of the appropriate selection, timing, and goals of fluid therapy is vital to optimize patient care.

Fluid Distribution and Movement

Water is the most abundant constituent of the body, comprising between 50% and 70% of total body weight. Variations in total body water (TBW) depend primarily on lean body mass, since fat and other tissue contain very little water (Table 47-1). Water is distributed within both intracellular fluid (ICF) and extracellular fluid (ECF) compartments. The distribution of water in an average adult male is shown in Table 47-2. The intracellular space contains two thirds of the TBW, with the remainder distributed to the extracellular space, which is further divided into interstitial and intravascular spaces in a 3:1 ratio. These fluid compartments are not contiguous, but may be treated as such due to similar composition and behavior.

Water freely crosses cell membranes. Osmotic forces within fluid compartments determine water distribution within the body. Intracellular and ECF environments are iso-osmolar, but physiochemically distinct due to tight regulation of dissolved solutes and proteins. Membrane-bound sodium–potassium-ATPase pumps compartmentalize sodium and potassium to the extracellular and intracellular spaces, respectively. Active restriction of sodium to the extracellular space is the foundation for isotonic sodium-based resuscitation solutions.

The intravascular fluid, or plasma, differs from all other fluid compartments in that it exists as a single continuous fluid collection and contains trapped protein moieties in a higher concentration than the surrounding interstitial fluid. These trapped proteins produce the colloid oncotic pressure (COP) that favors fluid flow into the vascular space. Fluid flux across vascular endothelial membranes is governed by Starling forces (Table 47-3). In health, transcapillary hydrostatic force is nearly opposed by COP. Small net losses from the vascular space are returned to the systemic circulation via the lymphatic system. Albumin typically accounts for 80% of COP, while large cellular moieties such as red cells and platelets contribute less oncotic pressure effect. Positive hydrostatic pressure, hypoalbuminemia, and pathologic endothelial permeability are common clinical conditions that enhance fluid extravasation from the vascular compartment. The clinical consequences may be large and persistent resuscitation requirements with the cost of cumulative tissue (e.g., lung, brain, gut) edema that can adversely impact function. Alteration of COP and enhanced retention of intravascular volume is one theoretical advantage of colloid-based fluids.

Effective Circulating Volume

Effective circulating volume (ECV) refers to the portion of intravascular volume contributing to organ perfusion. It falls with hypovolemia but does not necessarily correlate with volume status, since organ perfusion is also dependent on cardiac output (CO), arterial tone, and circulatory distribution. As an example, ECV may be compromised by limited CO despite optimized volume status.

The immediate consequence of hypovolemia is impaired oxygen delivery, which triggers a swift compensatory response. CO is the most important determinant of oxygen delivery, with the flexibility to compensate for reduced oxygen-carrying capacity and/or increased metabolic demands. In the setting of hypovolemia, the body acts to defend itself through adjustments to maintain perfusion pressure and oxygen delivery (Table 47-4).

At the macrocirculatory level, volume loss leads to decreased venous return and decreased CO. Reduced stretch sensed by aortic and carotid baroreceptors leads to swift sympathetic catecholamine release, resulting in peripheral vasoconstriction, tachycardia, and enhanced cardiac contractility. These compensatory measures attempt to maintain CO in the face of a falling stroke volume. Venoconstriction shunts blood from capacitance vessels and maintains intrathoracic blood volume and cardiac preload. Organ blood flow is directly proportional to perfusion pressure in most vascular beds, and vasoconstriction maintains critical arterial pressure. Preferential perfusion simultaneously shunts limited CO to vital organs, at the expense of reduced blood flow to noncritical (hepatosplanchnic, renal, cutaneous) organs. As such, mean arterial pressure (MAP) is maintained despite hypovolemia and organ hypoperfusion.

Signs and Symptoms

Hypovolemia primarily manifests as circulatory insufficiency. Signs and symptoms reflect organ dysfunction and the counterregulatory response set in motion to offset the hypovolemic state. Classically, hypovolemia is portrayed to follow a stepwise progression of signs and symptoms based on the volume deficit. The clinical reality is that signs of hypovolemia are highly variable depending on the culprit disease, acuity of evolution, and individual physiologic reserve. Compared with hemorrhage, sepsis presents a complicated hypovolemic state in which absolute fluid deficits are compounded by pathologic vasodilation and accelerated end-organ dysfunction. Children and healthy adults with vigorous compensatory mechanisms may tolerate large volume loss in the absence of severe clinical symptoms. In contrast, patients with limited cardiac reserve may poorly tolerate even minimal fluid loss.

Delayed capillary refill, dry axilla and mucus membranes, abnormal skin turgor, and sunken eyes are classic, but imperfect, hallmarks of hypovolemia. Symptoms of reduced CO such as fatigue, dyspnea, postural dizziness, or near syncope are common but are neither specific nor sensitive. Nonfocal confusion, agitation, and lassitude are common manifestations of hypovolemia in elderly patients.6 Organ dysfunction can be the heralding signal of hypovolemia and may occur in the absence of global hypoperfusion or hemodynamic instability. Oliguria, concentrated urine, and increased serum creatinine are examples. Electrolyte and acid–base derangements associated with hypovolemia may also produce a constellation of associated symptoms.

Blood Pressure

Shock defines a state of inadequate tissue perfusion in which tissue oxygen delivery is inadequate to meet metabolic needs. Contrary to popular belief, the term does not reflect perfusion pressure; shock may occur with low, normal, or elevated blood pressure. Inadequate perfusion in the setting of normal blood pressure is labeled compensated shock. Arterial pressure is intensely preserved via compensatory vasoconstriction, and normotension often masks clinical recognition of hypoperfusion and individual severity of illness. The difficulty in identifying these patients has spawned the terms occult hypoperfusion and cryptic shock to describe hemodynamically stable patients with microvascular insufficiency. Hyperlactatemia (>4 mmol/L) is an important clue to identify these patients.7

The majority of critically ill patients present in compensated shock with normal or near-normal blood pressures. Left unresuscitated, these patients may progress to frank hypotension. Brief episodes of hypotension are important markers of hypoperfusion and herald progressive hemodynamic deterioration.8,9 These self-limited hypotensive episodes represent progressive exhaustion of cardiovascular compensation and are the first sign of uncompensated shock.

Uncompensated shock, characterized by hypotension, is a late finding, and develops when physiologic attempts to maintain normal perfusion pressure are overwhelmed or exhausted. As such, hypotension should always be considered pathologic. MAP of less than 65 mm Hg, systolic blood pressure less than 90 mm Hg, and/or MAP more than 20 mm Hg below baseline should raise concern. Normal blood pressure values do not reliably indicate adequate oxygen delivery, and hypotension is typically found in the late stage of shock.10,11

It is important to understand the limitations of blood pressure measurements in critically ill patients. Automated blood pressure cuffs rely on the oscillometric method of blood pressure determination and may overestimate true arterial blood pressure in low-flow states.12,13 Direct auscultation with reliance on Korotkoff sounds may underestimate actual systolic blood pressure by as much as 30 mm Hg in low-flow states.14 The potential for large measurement errors in hemodynamically unstable patients warrants consideration for invasive arterial pressure monitoring.

Heart Rate

Sinus tachycardia is nonspecific, but should prompt careful clinical consideration of volume depletion, hemorrhage, and sepsis. Heart rate typically increases in the early stages of hypovolemia to maintain CO in the face of falling stroke volume. However, heart rate response to acute volume loss is highly variable. In otherwise healthy patients, volume loss of up to 20% fails to induce a tachycardic response.6 This compensatory response may be further blunted by comorbid disease and medications, especially β-blockers. Paradoxical and relative bradycardia occur in up to 30% of patients with traumatic and nontraumatic hemoperitoneum.15,16

Orthostatic Blood Pressure

The discriminative power of postural vital signs depends on appropriate testing and integration with specific clinical findings. Reassessment of supine resting blood pressure and pulse rate should be performed at least 2 minutes after standing, because all patients have a brief orthostatic response after standing. Postural pulse change greater than 30 beats/min is unusual in normovolemic patients.6 Severe postural dizziness with intolerance of the upright position confirms hypovolemia, in contrast to subjective symptoms that do not limit standing. Postural hypotension, defined as a systolic blood pressure decline of greater than 20 mm Hg, is seen in 10–30% of normovolemic patients. The postural hemodynamic response may also be altered due to aging and medications. Up to 30% of elderly patients demonstrate an orthostatic response in the absence of volume depletion.16

Shock Index

The shock index (SI) is the ratio of the heart rate to systolic blood pressure. The normal range for SI is 0.5–0.7. SI >0.9 may aid identification of patients with critical illness despite seemingly normal vital signs.17 SI identifies acute blood loss better than either HR or SBP alone.18 Unfortunately, elevated SI is not specific for hypovolemia and may be less accurate in clinical conditions with fever-associated tachycardia.

Indications

Circulatory failure is the final common pathway of many diseases and carries a wide differential diagnosis (Table 47-5). Inadequate circulating volume is the most common primary etiology of shock. Immunologically mediated pathologic vasodilation compounds the fluid deficit in many clinical conditions. Acute cardiac decompensation and pulmonary embolus are two exceptional situations in which limited volume resuscitation and priority of mechanical and catecholamine resuscitation are recommended.

Volume depletion describes the state of contracted ECF with clinical implications of compromised ECV, tissue perfusion, and function. It is distinguished from dehydration, which implies an intracellular water deficit characterized by plasma hypernatremia and hyperosmolarity. Hypovolemia may occur as a consequence of blood, electrolyte, and/or primary water loss (Table 47-6).

In any state of presumed or suspected volume depletion, rapid restoration of underlying fluid deficit is a first step to reverse hypoperfusion. Similarly, empirical volume therapy is a cornerstone of early resuscitative efforts in undifferentiated shock. Restoration of adequate oxygen delivery through fluid resuscitation initially relies on maximizing stroke volume. After the initial resuscitation and stabilization phase, ongoing fluid replacement can be more adequately tailored to specific clinical scenarios.

Intravenous Access

Appropriate intravenous access is vital to resuscitation. The determinants of flow through a rigid tube are shown in Table 47-7. The rate of volume infusion is determined by the dimension of the vascular catheter, and not by the size of the cannulated vein. Flow is directly proportional to the fourth power of the catheter radius and inversely proportional to the catheter length. Therefore, doubling the catheter size results in a 16-fold increase in flow, whereas doubling the cannula length decreases flow by half.

Central venous catheters (CVCs) enable hemodynamic monitoring and provide a reliable portal for volume therapy, vasoactive drug infusion, and serial blood sampling. Due to the differential catheter lengths, infusion rates through adult CVCs are up to 75% less than with peripheral catheters of equal diameter. In some circumstances, massive volume infusion may require the use of large-bore introducer (8.5–9.5 French) catheters that support flow rates approaching that of intravenous tubing at almost 1 L/min19 (Table 47-8). Additionally, manual compression of the fluid bag is an inefficient method of improving flow when compared with use of an external pressure bag.20

Endpoints of Resuscitation

Endpoints or markers of resuscitation are imperative to guide therapy during acute critical illness support (Table 47-9). Rapid restoration of perfusion pressure is a priority. Restoration of MAP to 60–65 mm Hg supports vital organ autoregulation.21 However, normalization of traditional markers of blood pressure, heart rate, and urine output does not guarantee adequate oxygen delivery or organ perfusion.1,10,11 Resuscitation aimed to these markers risks leaving the patient in persistent compensated shock.

Resuscitation aims to stabilize oxygen delivery to meet global and regional metabolic requirements. Central venous oxygen saturation (ScvO2) and serum lactate have emerged as rapid, reliable global perfusion markers. ScvO2 reflects the systemic balance of oxygen delivery and utilization. Decreasing oxygen delivery is compensated by increased tissue oxygen extraction, resulting in a fall in ScvO2 below the normal 70%. ScvO2 is a practical bedside measurement that is sampled from a catheter (CVC or PICC) positioned in the superior vena cava. ScvO2 response is rapid and dynamic, such that monitoring provides immediate feedback on resuscitation efforts (or clinical deterioration).

Admission lactate and base deficit (BD) predict morbidity and mortality independent of hemodynamics.7,22–24 These markers of illness severity are conversely useful as endpoints of resuscitation. Rapid lactate clearance is associated with improved outcome from critical illness and should be incorporated in the goals of resuscitation.25–28 It is important to recognize that initial BD often correlates with serum lactate, but initial and serial measures may be confounded by underlying disease (e.g., renal insufficiency, malnutrition), resuscitation fluid (e.g., normal saline [NS]–induced acidosis), and other therapies (e.g., bicarbonate, blood products).

The optimal endpoint of resuscitation remains controversial. Novel markers such as tissue capnometry, oximetry, and near-infrared spectroscopy hold promise but their roles remain to be clarified. However, we cannot expect a single resuscitation endpoint to perform in all clinical circumstances. As such, a multimodal approach seeking to normalize a combination of both global and regional perfusion markers is most prudent (Table 47-9).

The empirical volume challenge remains the standard means of early fluid resuscitation. Volume expansion is achieved by infusing serial aliquots of isotonic fluid under direct observation and is appropriate in acute undifferentiated shock or when there is obvious or suspected hypovolemia. The use of crystalloid (10–20 mL/kg) or colloid (5–10 mL/kg) is infused quickly over 15–20 minutes, and serial boluses are titrated to the clinical endpoint objective while monitoring for adverse effects. A positive clinical response to volume loading confirms volume responsiveness but does not predict further response to therapy. This can contribute to overly aggressive volume expansion.

Total volume requirements are difficult to predict at the onset of resuscitation and are often underestimated. Classic hypovolemia that occurs with acute hemorrhage or fluid loss may stabilize rapidly with appropriate volume expansion. The 3:1 rule of hemorrhage resuscitation suggests that 3 volumetric unit of crystalloid are required to replete the ECF deficit of 1 unit of blood loss. However, experimental models confirm the experience in severely traumatized patients whose fluid requirements exceed the 3:1 suggestion.29 Pathologic vasodilation and transcapillary leak contribute to the need for ongoing volume replacement. Crystalloid requirements average 40–60 mL/kg in the first hour of septic shock but may be as high as 200 mL/kg to normalize perfusion parameters.30–32

The ability of fluid administration to improve stroke volume depends on a number of variables that include venous tone and ventricular function. Following initial resuscitation, the ability of fluid to further improve macrocirculatory flow is as low as 50%.33,34 Volume or preload responsiveness refers to the ability to augment stroke volume with fluid administration. In contrast to the empirical volume challenge, volume responsiveness is gauged prior to fluid administration with the information used to guide whether fluid administration is part of the solution to reverse clinical hypoperfusion. Fluid loading in nonresponsive patients should be avoided because it delays appropriate therapy and contributes to volume overload and organ dysfunction, including hypoxemic respiratory failure and abdominal compartment syndrome.4,5 Recent trials highlight the importance of fluid balance in organ function and morbidity.2

Predicting Volume Responsiveness

Invasive hemodynamic measurements are frequently used as surrogates of preload and predictors of volume responsiveness. Central venous pressure (CVP) monitoring is widely advocated. In the absence of conflicting data, a target CVP of 8–12 mm Hg is recommended to optimize preload prior to the institution of pressor and inotropic support.1,31

Unfortunately, cardiac pressure surrogates of preload (CVP and PAOP) reflect the net influences of intravascular volume, venous tone, cardiac function, and intrathoracic pressure. These myriad influences confound their ability to reflect intravascular volume status or preload responsiveness of an individual patient.33,34 There is no consistent threshold CVP to reliably estimate response to fluid administration.35,36 Values that are considered low, normal, or high can be found in patients who respond positively to fluid. Obstructive lung disease, positive pressure ventilation, myocardial dysfunction, reflex venoconstriction, and erroneous measurements are several examples that can result in elevated CVP in a volume-responsive patient. CVP rise coupled to clinical improvement with volume loading corroborates fluid responsiveness, but does not anticipate further effect.

Volumetric measures of preload including stoke volume, right and global end-diastolic volume, and left ventricular end-diastolic area can be obtained with several monitoring techniques. These volumetric surrogates of preload are intuitively more desirable, but they too have limited predictive value because discriminatory thresholds are imprecise and infrequent in clinical practice.33 Serial volumetric data in response to therapy may assist in individual patient management, but the dynamic nature of cardiovascular function during critical illness may confound data interpretation.

Dynamic Indices of Fluid Responsiveness

Fluid responsiveness is best predicted by dynamic indices of preload reserve. Respirophasic variation in stroke volume during positive pressure mechanical ventilation is among the most reliable signs of preload responsiveness.33,34 Positive pressure ventilation induces cyclic alteration in preload. A resulting variation of systolic pressure, pulse pressure, and stroke volume greater than 13% identifies patients capable of augmenting stroke volume in response to fluid administration. A regular (preferably sinus) rhythm, positive pressure ventilation on >8 mL/kg tidal volume, and absence of significant patient interaction with the ventilator are important requirements for interpretation of these data.

Passive leg raising (PLR) is a provocative maneuver that tests whether a reversible volume challenge results in improved stroke volume.37,38 This is an attractive option, since it provides immediate information to guide therapy without the administration of potentially unnecessary fluid. PLR results in the translocation of venous blood from the lower extremities to the thorax. The transient increase in preload improves stroke volume within minutes. Rapid feedback stroke volume measurement tools are required to identify the brief response to PLR. Sensitivity and specificity of PLR in predicting volume responsiveness is greater than 95% in a wide variety of patients, including ventilated and spontaneously breathing patients, and in those with irregular cardiac rhythms.39

The goal of fluid resuscitation is intravascular expansion to optimize stroke volume. Early resuscitation and ongoing replacement of fluid deficits may be performed using a variety of fluid choices. Each possesses specific benefits and potential disadvantages in given clinical scenarios, and an understanding of fluid composition is important (Tables 47-10 and 47-11).

Crystalloid

Isotonic sodium-based crystalloids preferentially distribute to the extracellular compartment, which includes the vascular space. One-liter infusion of isotonic crystalloid distributes approximately one quarter into the vascular compartment. This is the basis for the 3:1 rule often cited for resuscitation in acute hemorrhagic shock. The ratio more closely approximates 7:1 or 10:1 in severe hemorrhage due to decreased COP secondary to hemorrhage, capillary leak, and crystalloid replacement. Interstitial tissue edema is the cost of such high-volume crystalloid requirement.

Fluid selection appears less important than volume dosage titrated to an appropriate therapeutic endpoint. NS (0.9%) and lactated Ringer's (LR) are the two most commonly used isotonic resuscitation solutions. Evidence of clinical superiority for either is lacking. However, the source of hypovolemia, associated electrolyte derangements, and volume requirements should impact fluid selection.

NS supplies a supraphysiologic sodium and chloride load that should be expected to induce hyperchloremic metabolic acidosis when administered in large volumes. This may be advantageous to correct volume and electrolyte disturbances in cases of metabolic alkalosis such as loss of gastric secretions (e.g., vomiting, gastric outlet obstruction, NG suctioning).

LR, or Hartmann's solution, was originally introduced in the 1930s by adding sodium lactate as a buffer to Ringer's solution for the treatment of metabolic acidosis. It is a more physiologic fluid, containing potassium and calcium in concentrations near plasma levels. Due to its more physiologic pH, it is preferred in large volume resuscitation. One caution is that calcium within the solution can bind to medications and the citrated blood anticoagulant, such that it is not a compatible transfusion fluid.

Colloids

Colloid solutions are composed of electrolyte preparations fortified with large-molecular-weight molecules (MW >30,000). The presence of these large molecules contributes to total oncotic pressure, which favors retention of fluid within the vascular space. The ideal colloid solution has an oncotic pressure similar to plasma, which permits replacement of the plasma volume without distribution to other fluid compartments. The net effect and theoretical benefit of colloid infusion is intravascular expansion without accompanying expansion of the interstitial compartment. The potency of colloid solutions on plasma expansion differs with individual fluids. Higher COP provides greater expansion of the plasma volume. Albumin, dextran, and blood are naturally occurring colloids, while synthetic colloids include modified gelatins, hydroxyethyl starch (HES), and hemoglobin solutions. Albumin is the only colloid that contains molecules of uniform weight. Other colloid solutions are comprised of polymers with a wide variety of molecular sizes. However, the average molecular weight of a colloid solution is an unreliable indicator of intravascular persistence and molecular weight distribution curves provide the best indicator of intravascular effect.

Albumin

Human albumin is a single polypeptide solution derived from pooled human serum albumin, and is available in 5% and 25% concentrations. Five percent albumin is iso-oncotic to plasma with greater than 70% of infused volume retained within the vascular space. It is recommended for resuscitation of patients with severe hypoalbuminemia and cirrhosis.31,40,41 Hyperoncotic albumin (25% albumin) was initially developed in the 1940s for combat resuscitation. Infusions of hyperoncotic albumin result in vascular expansion greater than two times the administered volume.42 Besides the obvious benefits of small volume resuscitation, improved portability, and more rapid hemodynamic stabilization, hyperoncotic albumin has additional advantages. Synergistic interaction with administered drugs and primary antioxidant effects are hypothesized explanations for the improved morbidity and mortality linked to hyperoncotic albumin for complicated hypoalbuminemic states including decompensated end-stage liver disease.43,44 Increased COP mobilizes interstitial edema, and the effects of hyperoncotic albumin are relatively long-lasting, persisting for up to 12 hours after infusion. As such, hyperoncotic albumin is often matched with loop diuretic therapy to mobilize fluid in volume-overloaded patients.45

Hydroxyethyl Starch

HES solutions are semisynthetic polymers of nonuniform size derived from amylopectin. HES is typically available as a 6% isotonic solution. Intravascular persistence is maximized by substitution of hydroxyethyl groups, which limit degradation by amylase. Smaller polymers of less than 60 kDa are rapidly eliminated through glomerular filtration, while medium- and large-molecular-weight polymers are cleared by the reticuloendothelial system. High-molecular-weight HES reduce factor VIII and von Willebrand factor that may lead to coagulopathy. Medium- and small-molecular-weight polymers have less effect on coagulation. The overall effect of a 6% HES solution is volume expansion comparable to 5% albumin. Renal dysfunction and coagulopathy that complicated early generation synthetic colloids do not appear clinically significant with new-generation HES solutions.

Gelatin Solutions

Gelatin polypeptides are derived from bovine collagen. Urea-bridged gelatin is derived by cross-linking polypeptides derived from cattle bone, while succinylated gelatin is produced by thermal degradation of calf skin collagen. Approximately 80% of urea-bridged gelatin is smaller than 20 kDa and is therefore rapidly excreted by the kidneys, with a short intravascular persistence (2–3 hours). Gelatin solutions are generally considered to have no effect on clotting. Since gelatins are formed from degraded animal collagens, there is an inherent risk of anaphylaxis.

Dextrans

Dextran solutions are composed of polysaccharides of varying molecular weights. Currently available solutions are 6% dextran 70 and 10% dextran 40. Dextran 40 is hyperoncotic and therefore expands intravascular volume by more than the treatment infused. Solutions reduce blood viscosity and enhance fibrinolysis. The majority of dextran is excreted by the kidney, although dextran 40 has been associated with kidney injury, especially in the presence of preexisting renal dysfunction or hypovolemia.

Colloid Resuscitation

There is no clear mortality impact of colloid resuscitation over crystalloid use across a broad range of critical illness.46,47 However, strategic use may provide advantage in specific situations, as part of a mixed resuscitation or in attempt to avoid the risk associated with large volume crystalloid resuscitation. Vascular retention of colloids makes them efficient volume expanders. Although equally effective when titrated to the same clinical endpoints, crystalloid solutions require two to four times more volume for equivalent resuscitation.48 Colloids therefore restore intravascular volume and tissue perfusion more rapidly if access and administration rate are limited. Furthermore, dilutional hypoalbuminemia, transcapillary fluid shift, and interstitial and pulmonary edema are limited. Albumin improves organ function and morbidity and is superior to crystalloids for intravascular volume expansion during hemodialysis, following large-volume paracentesis and in combination with antibiotic therapy for spontaneous bacterial peritonitis.41,44,49,50 Traumatic brain injury remains one important exception in which isotonic albumin is associated with increased risk of adverse outcome compared with crystalloid resuscitation.46,51

Hypertonic Saline

Hypertonic sodium (HS) solutions, with sodium concentrations ranging from 3% to 7.5%, rapidly expand intravascular volume by mobilizing water from interstitial and intracellular spaces. Small infusions expand plasma by several times the infused volume without resultant expansion of the interstitial fluid space and edema seen with crystalloid infusions.52 Additional benefits include improved cardiovascular performance secondary to positive inotropic effects and microvascular vasodilation, improved microcirculatory flow, and attenuation of the inflammatory response. This combination of effects is attractive for use in volume expansion in trauma and septic patients alike. Used alone, however, the hemodynamic effects of hypertonic crystalloid are transient. HS is generally used in combination with hyperoncotic colloid (6% dextran or 10% hetastarch). HS is safe, but there are insufficient data to conclude that hypertonic saline is better than isotonic crystalloid for the resuscitation of patients with burns, trauma, or sepsis. Multitrauma patients with traumatic brain injury remain the most common indication for HS, but outcome benefit also remains equivocal in this group.53,54

Minimal Volume Resuscitation of Hemorrhagic Shock

Hemorrhagic shock poses a unique challenge between balancing the timing and type of resuscitation in relation to the achievement of hemostasis. On one hand, hypotensive patients should be stabilized with rapid fluid infusion to maintain perfusion to essential organs. However, overly aggressive fluid resuscitation before control of bleeding may result in increased blood loss and mortality.55 Factors that prevent hemostatic plug formation and allow renewed bleeding such as increased volume and blood pressure, decreased blood viscosity, and dilution of clotting factors are all associated with fluid resuscitation.

Strategic, limited volume resuscitation reemerged in the 1980s as the value of early prehospital resuscitation in penetrating trauma was questioned. A prospective trial comparing immediate and delayed fluid resuscitation in hypotensive patients with penetrating torso injuries showed improved mortality, fewer complications, and a shorter hospital length of stay with delayed resuscitation.56,57 Limited prehospital resuscitation with judicious use of fluids may offer the optimal approach with conventional resuscitation ensuing after surgical hemostasis is achieved.58 The degree and duration of permissive hypotension remains unclear, although current recommendations target SBP of 70 mm Hg. Patients with concomitant brain injury are not candidates for this strategy (Table 47-12).

Burn Resuscitation

Patients with second- and third-degree burns exhibit marked fluid shifts related to denuded skin, injured tissue, and systemic inflammatory response. Aggressive fluid resuscitation is necessary to restore intravascular volume and maintain end-organ perfusion. Early anticipation of these large fluid requirements prevents underresuscitation. Initial fluid requirements are most commonly calculated according to the Parkland formula (Table 47-13).

Formula calculations are based on the time of injury as opposed to the time to medical attention, and should incorporate prehospital fluid administration. LR is the preferred crystalloid solution. Several formulas exist but no single method is clearly superior.59 All formulas are intended to provide an initial guide for resuscitation requirements. Actual fluid needs may vary significantly, necessitating modifications based on individual status.60 Strict adherence to a calculated goal may result in overresuscitation or underresuscitation. Overresuscitation is common and contributes to increased pulmonary complications and morbidity. Maintenance fluid requirements should be allocated in addition to burn formula replacement. Urine output greater than 1 mL/kg/h is a traditional endpoint of acute burn resuscitation and may be augmented by perfusion endpoints discussed above.

In contrast to resuscitation therapy, the goal of maintenance fluid therapy is normal body fluid composition and volume. Fluid orders anticipate daily fluid requirement, ongoing losses, and coexisting electrolyte abnormalities. Although often ordered concurrently, daily physiologic fluid estimate (true maintenance) should be consciously distinguished from therapy aimed to slowly replace an existing fluid deficit.

Routine water and electrolyte maintenance are based on normal energy expenditure, sensible loss from urine and stool, and insensible loss from the respiratory tract and skin. Calculations assume euvolemia and are adjusted for body mass (Table 47-14). Greater per kilogram fluid requirements in children are proportionate to TBW and metabolism. All maintenance prescriptions should be individualized; energy expenditure, fluid losses, and electrolyte status vary with disease and dictate rate and electrolyte modifications. For example, exfoliative skin disease, increased work of breathing, and fever enhance insensible loss. Measurable nasogastric, fistula, ostomy, and urinary drainage can be estimated or replaced by drainage volume. Limitation of fluid and potassium is an important disease-specific modification for patients with renal insufficiency.

Hypotonic solutions with or without dextrose and potassium are popular fixed-combination maintenance solutions. Hospitalized patients often suffer impaired free water excretion due to nonosmotic antidiuretic hormone (ADH) release, making them vulnerable to hyponatremia. Serum sodium concentration provides a simple and accurate marker of hydration status. Isotonic maintenance solutions should be considered in patients (including children), especially those with serum sodium <138 mEq/L.61–63 Glucose infusions are best formulated by adding dextrose to an electrolyte solution (e.g., LR, NS, 0.45 NS) rather than using 5% dextrose (D5W), which behaves as electrolyte free water on sugar metabolism.

1. The critical window to reverse organ hypoperfusion is measured in hours, emphasizing the need for rapid recognition and correction of shock.

2. The majority of ED patients who require resuscitation present in compensated shock with normal blood pressure.

3. Early recognition of circulatory insufficiency must be coupled with timely resuscitation to impact patients.

4. Normalization of vital signs does not insure adequate systemic perfusion or completion of resuscitation.

5. The clinical endpoint used to guide dosage of fluid resuscitation is more important than the individual product (i.e., crystalloid vs. colloid) selection.

6. Overly aggressive fluid resuscitation and positive fluid balance negatively impact patient morbidity.

7. Dynamic markers of volume responsiveness are important guides for fluid therapy.