Intravascular Volume Replacement Therapy


This chapter describes conventional concepts about body fluid spaces. It carefully describes physiological principles about fluid shifting in the body and responses to fluid challenges. It further describes different monitoring principles on how to estimate fluid volumes. Furthermore it gives a body fluid dynaimic theory in a mathematical way. Finally, it gives recommendations on how to handle fluid requirements in daily practice

Key words

body water, fluid spaces, fluid challenge, monitoring of fluid status, body fluid dynamics and volume kinetics, guiding principles for fluid


Historical Perspective

Plasma volume replacement is important in the perioperative period. The body and cardiovascular system are exposed to many challenges, such as neurohumoral adaptations, evaporation, fluid redistribution, and blood loss, that necessitate interventions. To achieve this, fluids are administered intravenously following protocols based on tradition, expert recommendations, and often limited evidence. There is an ongoing debate concerning the ideal composition and amount of intravenous fluids necessary for perioperative management. In the past century, recommendations varied from fluid restriction to giving liberal amounts for resuscitation. Surprisingly, intravenous (IV) fluids have been regarded as rather harmless, resulting in nothing worse than volume overload, which is often viewed as a minor problem. However, IV fluids can be deleterious in large amounts if not timed according to patient needs. In the late 1970s and early 1980s, acute lung injury caused by increased filtration rate across pulmonary capillaries and subsequent pulmonary inflammation was suggested as a plausible consequence. It has never been proven that large amounts of crystalloids cause acute respiratory distress syndrome; however, the link to abdominal compartment syndrome, which can be the result of large-volume crystalloid resuscitation, is compelling. Traditionally, the rationale supporting liberal perioperative fluid administration included assumptions that preoperative fasting resulted in hypovolemia, that insensible losses increased considerably during surgery, and that some fluid was distributed to a “third space.” Any resulting fluid overload was considered harmless because the kidneys had the capacity to eliminate the excess.

The primary problem the clinician faces is that individual hydration and volume states are unknown before surgery. There are a few simple tests or physical maneuvers that can assess reliably either the level of hydration or the intravascular volume status. The clinician usually must rely on indirect nonspecific clinical signs to estimate the volume status of the cardiovascular system. Because the vascular system is highly reactive to neurohumoral changes and positioning, predicting the disposition of IV fluids is difficult.

Very few healthy stable surgical patients admitted for elective minor operations require significant amounts of fluid, and thus the perioperative fluid management of these patients is straightforward. In contrast, fluid management of critically ill patients can be extremely demanding, and timing is very important. These patients sometimes require volume support, usually with extensive monitoring to guide fluid resuscitation. However, evidence suggests that several currently used monitors have limitations in measuring the adequacy of intravascular volume.

Conventional Concepts

Body Water

With respect to traditional fluid therapy, the body is conceptualized as an interconnected group of anatomic spaces among which fluids distribute. However, this static concept hardly reflects the complexity of how fluids dynamically distribute over time. Total body water (TBW) is the amount of sodium-free water in the whole body, commonly divided into the extracellular fluid (ECF) space and the intracellular fluid (ICF) space. Under normal conditions, ECF constitutes 20% of total body weight and ICF 40% of TBW for an adult.

The ECF consists of the plasma volume and interstitial fluid volume ( Fig. 41.1 ). The plasma volume (PV) is relatively small, which is important to understand when boluses of fluids are given. The interstitial fluid volume contains water but is mainly bound by a gel-like composition of proteoglycan filaments and collagen fibers. Infused sodium ion (Na + ) distributes mainly within ECF, which contains equal Na + concentrations in PV and interstitial fluid volume (approximately 140 mM). The predominant intracellular cation K + (potassium ion) has an intracellular concentration of approximately 150 mM.

Fig. 41.1

A static schematic of the distribution of body water. Water constitutes about 60% of body weight and is unequally distributed between extracellular and intracellular spaces. BW, Body weight in kilograms; ECF, extracellular fluid; ICF, intracellular fluid; IFV, interstitial fluid volume; PV, plasma volume; TBW, total body water.

Measurement of Body Fluid Spaces

There are several methods to measure the various fluid spaces. TBW can be estimated by anthropometric formulae, commonly validated from tracer dilution methods. Anthropometric predictions of various physiologic properties depend on height, weight, age, gender, and race; these population models naturally result in various degrees of inaccuracy when applied to individuals. For example, TBW in liters can be estimated by the formulas :

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='Male TBW=2.447−(0.09156×age)+(0.1074×height)+(0.3362×weight)’>Male TBW=2.447(0.09156×age)+(0.1074×height)+(0.3362×weight)Male TBW=2.447−(0.09156×age)+(0.1074×height)+(0.3362×weight)
Male TBW = 2.447 − ( 0.09156 × age ) + ( 0.1074 × height ) + ( 0.3362 × weight )

<SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='Female TBW=−2.097+(0.1069×height)+(0.2466×weight)’>Female TBW=2.097+(0.1069×height)+(0.2466×weight)Female TBW=−2.097+(0.1069×height)+(0.2466×weight)
Female TBW = − 2.097 + ( 0.1069 × height ) + ( 0.2466 × weight )

Inaccuracy when applied to the individual can be substantial. This equation does not take into account that TBW decreases with age. The amount of body fat also influences TBW; fat varies inversely with water. TBW can be measured by isotope dilution techniques, preferably with nonradioactive isotopes such as the stable water isotopes 2 H 2 O and H 2 18 O, and the precision can be as high as 1%. However, these methods are not clinically feasible owing to the complex experimental setup and mixing time required for the tracer. Another technique of estimating TBW is by bioelectrical impedance analysis. Although simple, quick, and cheap, there are many potential sources of error.

ECF can also be estimated by tracer techniques. The most commonly used tracer is the bromide anion (Br ). Two drawbacks of using bromide as a dilution tracer are that it requires a long mixing time and does not distribute equally through the ECF. The blood volume is commonly predicted by anthropometric formulae but individual variation can be considerable. Because the blood volume includes erythrocytes (in ICF) and the PV, the two components can be estimated individually. The volume of erythrocytes can be determined by isotope labeling with chromium 51 ( 51 Cr) and technetium 99 ( 99 Tc), for example, but there are other nonradioactive methods such as labeling erythrocytes with fluorescein. Another methods uses a semiautomated blood volume analyzer that can provide rapid and fairly accurate information when assessing blood volume in blood loss situations and making critical transfusion decisions. It uses albumin–iodine 131 ( 131 I) and results are available in 20 minutes.

Maintenance Requirements for Water, Sodium, and Potassium

TBW content is regulated by the intake and output of water. Water intake includes ingested liquids plus an average of 750 mL ingested in solid food and 350 mL generated metabolically. Perspiration losses are approximately 1000 mL/day and gastrointestinal losses are about 200 mL/day ( Fig. 41.2 ).

Fig. 41.2

Schematic of daily fluids turnover. Arrow direction indicates intake or output. Water intake is required to offset gastrointestinal, urinary and insensible losses.

Thirst is the primary mechanism of controlling water intake and is triggered by an increase in plasma tonicity or by a decrease in ECV (see Chapter 42 ). Reabsorption of filtered water and Na + is enhanced by changes mediated by the hormonal factors antidiuretic hormone (ADH), atrial natriuretic peptide (ANP), and aldosterone (see Chapters 40 and 42 ). Renal water handling has three important components: (1) delivery of tubular fluid to the diluting segments of the nephron, (2) separation of solute and water in the diluting segment, and (3) variable reabsorption of water in the collecting ducts. In the descending loop of Henley, water is reabsorbed while solute is retained to achieve a final osmolality of tubular fluid of ~1200 mOsm/kg. This concentrated fluid is then diluted by active reabsorption of electrolytes in the ascending limb of the loop of Henle and distal tubule, both of which are relatively impermeable to water. As fluid exits the distal tubule and enters the collecting duct, osmolality is ~50 mOsm/kg. Within the collecting duct water reabsorption is modulated by ADH (also called vasopressin). Vasopressin binds to V2 receptors along the basolateral membrane of the collecting duct cells and stimulates synthesis and insertion of aquaporin-2 water channels into the luminal membrane of collecting duct cells to facilitate water permeability.

Plasma hypotonicity suppresses ADH release, resulting in excretion of dilute urine. Hypertonicity stimulates ADH secretion, which increases the permeability of the collecting duct to water and enhances water reabsorption. In response to changing plasma Na + , differences in the secretion of ADH can change urinary osmolality from 50 to 1200 mOsm/kg and urinary volume from 0.4 to 20 L/day. Other factors that stimulate ADH secretion, though none as powerfully as plasma tonicity, include hypotension, hypovolemia, and nonosmotic stimuli such as nausea, pain, and medications, including opioids.

Two powerful hormonal systems regulate total body Na + . The natriuretic peptides, ANP, brain natriuretic peptide, and C-type natriuretic peptide, defend against Na + overload and the renin-angiotensin-aldosterone axis defends against Na + depletion and hypovolemia. ANP, released from the cardiac atria in response to increased atrial stretch, produces vasodilation and increases the renal excretion of Na + and water. ANP secretion is decreased during hypovolemia. Even in patients with chronic (nonoliguric) renal insufficiency, infusion of ANP in low doses (i.e., that do not produce hypotension) increases Na + excretion and augments urinary losses of retained solutes. Aldosterone is the final common pathway in a complex response to decreased effective blood volume, whether decreased effective volume is true or relative (e.g., edematous states or hypoalbuminemia). In this pathway, decreased stretch in the baroreceptors of the aortic arch and carotid body and stretch receptors in the great veins, pulmonary vasculature, and atria result in increased sympathetic tone. Increased sympathetic tone, in combination with decreased renal perfusion, leads to renin release and formation of angiotensin I from angiotensinogen. Angiotensin-converting enzyme converts angiotensin I to angiotensin II, which stimulates the adrenal cortex to synthesize and release aldosterone. Acting primarily in the distal tubules, high concentrations of aldosterone cause Na + reabsorption and can reduce urinary excretion of Na + nearly to zero. Intrarenal physical factors are also important in regulating Na + balance. Sodium loading decreases colloid osmotic pressure, thereby increasing the glomerular filtration rate (GFR), decreasing net Na + reabsorption, and increasing distal Na + delivery, which in turn suppresses renin secretion.

In healthy adults, sufficient water is required to offset gastrointestinal losses of 50 to 200 mL/day, insensible losses of 850 to 1200 mL/day (half of which is respiratory and half cutaneous), and urinary losses of about 1000 mL/day (see Fig. 41.2 ). Urinary losses exceeding 1000 mL/day can represent an appropriate physiologic response to ECV expansion or an inability to conserve salt or water.

Daily requirements for Na + and K + are approximately 75 mEq/day and 40 mEq/day, respectively, although wider ranges of Na + intake than K + intake are physiologically tolerated because conservation and excretion of Na + are more efficient than of K + . Therefore healthy 70-kg adults require 2500 mL/day of water containing a Na + of 30 mM and a K + of 15 to 20 mM. Intraoperatively, fluids containing Na + -free water (i.e., Na + <130 mM) are rarely used in adults because of the necessity for replacing isotonic losses and the risk of postoperative hyponatremia, particularly in children.


During surgery and trauma, fluid extravasation is enhanced: capillary permeability increases and permits more fluid to escape to the interstitium. The ECV consists of the plasma (approximately 3.5 L in a 70-kg man), the interstitial space (approximately 10.5 L), and small amounts of transcellular fluids, such as gastrointestinal fluids, cerebrospinal fluid, and ocular fluid. The interstitium is a mixture of fluid and fibrillar structures. The main nonfluid components are fibrils and the interstitial ground substance that can be subdivided into a colloid-rich and a water-rich phase. Fibrils are mainly collagenous, reticular, and elastic. The amorphous ground substance or gel-like matrix is produced by the same cell types as the fibrillar components. It contains several different glycosaminoglycans (mucopolysaccharides). Plasma proteins passing the capillary wall are mainly restricted to a random network of interstitial channels corresponding to colloid-poor, water-rich areas. Edema causes and increases hydration and depolymerization of the ground substance ( Fig. 41.3 ).

Fig. 41.3

A capillary vessel with surrounding tissue. Endothelial cells permit leakage of water, small solutes, and proteins. Conditions can differ owing to pathology. Abbreviations refer to the Starling formula, which is modified by the existence of the glycocalyx. The inner lining of the glycocalyx and trapped plasma together form the endothelial surface layer (ESL). A revised Starling formula would be as follows: J v = K f ( P v P i − σΠ esl − Π s ). Within the ESL there is practically no circulation. GAGs, J v , Net filtration; K f , filtration coefficient; P i , interstitial hydrostatic pressure; P v , capillary hydrostatic pressure; Π i , interstitial oncotic pressure; Π esl , oncotic pressure within the endothelial surface layer; Π s , oncotic pressure beneath the endothelial surface layer; Π v , venous vascular oncotic pressure; σ , reflection coefficient;

(Modified from Chappell D, Jacob M, Hofmann-Kiefer K, et al. A rational approach to perioperative fluid management. Anesthesiology. 2008;109:723–740.)

Fluid distribution within the human body is related to the distribution of osmotically active substances. The physiologic distribution is maintained by biologic barriers and adenosine triphosphate–requiring pumps. The vascular wall is impermeable to larger molecules or proteins, with normal fluid distribution subject to an intact inner lining of the endothelial wall (glycocalyx). When outward and inward pressures are accounted for, there is a small net filtration of fluid. However, deterioration of the glycocalyx, as seen in systemic inflammatory conditions, is of greater importance than interstitial protein concentration for fluid escape through the endothelial barrier. The endothelial glycocalyx consists of a variety of transmembrane and membrane-bound molecules. These are mainly syndecans and membrane-bound glypicans that contain heparan sulfate and chondroitin sulfate chains. The thickness of the glycocalyx is about 1 µm. Together with some plasma proteins that are membrane bound, hyaluronan, and dissolved glycosaminoglycans, the inner endothelial surface layer is tens of nanometers thick. Degradation of the glycocalyx leads inevitably to increased capillary leakage and interstitial edema, which is strongly correlated with a decrease in tissue oxygenation. Reduction of the endothelial glycocalyx, such as in sepsis, major vascular surgery with global or regional ischemia, or diabetes mellitus, can lead to changes such as increased permeability and release of proteases, tumor necrosis factor α, oxidized low-density lipoprotein, and atrial natriuretic peptide. Even the shear force of IV solutions infused too rapidly have been suggested to cause deterioration of the glycocalyx.

Smaller molecules such as crystalloids are probably not much affected because they can escape through the barrier anyway. Filtration normally leads to only a moderate shift and accumulation of fluid because both increased lymph flow and interstitial hydrostatic pressure, in combination with dilution of interstitial protein, limit excess accumulation of interstitial fluid. There is also normal transcapillary escape of albumin of about 5% of the intravascular albumin per hour, which is increased in a variety of diseases. Transfer of albumin back to the intravascular space occurs via the lymph, but can be impeded by surgery and inflammation.

Third Space

Clinicians used to replace losses to the third space during surgical procedures. The concept of the third space has been a topic of debate and controversy for decades. Loss to the interstitial space is an inevitable phenomenon that consists mainly of small molecules leaving through an intact vascular barrier. When this loss caused by trauma and surgery becomes pathologic, it can be considered an accumulation within the interstitial space or the “functional” ECV. The accumulated fluid is normally removed by the lymphatic system as described earlier. However, when there is an overload, either by excessive fluid load or cessation of urinary production, the result is an even more pronounced accumulation of interstitial fluid. Fluid within the hypothetical third space is thought to be a “nonfunctional” body fluid separated from the interstitial space. Examples of third space would be fluid in the peritoneal cavity, bowel, and traumatized tissues, with the key point being that losses to these spaces are no longer in rapid equilibrium with the ECF and thus are nonfunctional.

A major problem is that the third space cannot be measured, and there is considerable controversy regarding its existence. It is common belief that fluid spaces can be identified by applying a tracer intravenously. The tracer is assumed to equilibrate within the desired space, and the dye or isotope enables identification of the distribution volume. Unfortunately, interpretation of tracer kinetic trials is not simple. The volume measured is the volume of distribution of a tracer and might not represent a volume of any clinical relevance. Furthermore, tracers are subject to differences in perfusion and equilibration time. For example, Na + is rapidly transported to the intracellular compartment, bromide enters red blood cells and is excreted in bile, and the sulphate tracer 35 SO 4 is bound to plasma components and can accumulate in the liver and kidneys (and during shock in muscular tissue as well). If the time for equilibration is too short, the concentration of any tracer in blood can be incorrectly high and the calculated volume of distribution will appear contracted. To overcome this, tracer spaces should be calculated from multiple blood samples with continued sampling until equilibrium is demonstrated.

The theory behind the third space concept is mainly based on studies by Shires and coworkers in the early 1960s performed on both animals and humans using 35 SO 4 to identify the functional ECV. This tracer has a short equilibration time and the extracellular space can thus be calculated from a single or just a few blood samples. Using this sulphate tracer it was possible to identify an apparently contracted ECF space, which led these investigators to believe this space needed replacement by crystalloids. The Shires group had considerable impact on researchers and clinicians in the 1960s, but few other researchers have managed to replicate their findings. In a study of 50 young male war casualties, ECV was estimated by measuring the 35 SO 4 concentration in multiple blood samples. Comparison of one group with minor injuries with a control group of base camp subjects revealed a general state of dehydration of the soldiers in the field. In another study, when corrected for the capillary refill phenomenon, comparison of the three shock groups with the control group of combat subjects showed that the ECV changed in accordance with calculated fluid balance. Other investigators studied 10 healthy young male volunteers who lost 11% to 14% of their blood volume with or without subsequent replacement with lactated Ringer (LR) solution. None of the subjects developed shock, and the ECV changed in accordance with the fluid balance.

Only trials using the sulphate tracer and short equilibration times have reported a “third space contraction.” Several researchers who were using tracers such as Na + or bromide all reported ECV expansion during surgical conditions. Trials calculating the ECV from multiple blood samples have found the ECV either unchanged or expanded after surgery.

The concept of a loss to third space needing replacement was introduced simultaneously with the conflict in Vietnam. Liberal fluid administration together with improved evacuation and field surgical care (and improvement in other treatment modalities) might have saved many lives. Renal failure, which had been a major problem in World War II and the Korean conflict, was almost unheard of. On the other hand, “the wet lung syndrome” (later diagnosed as acute respiratory distress syndrome) was reported in otherwise healthy soldiers resuscitated with large amounts of fluids after traumatic injury. However, the evidence for respiratory distress caused by fluid overload is not compelling.

Abdominal compartment syndrome is, however, linked to overzealous crystalloid administration. In a prospective randomized trial of patients undergoing colorectal surgery, a restricted IV fluid regimen omitting replacement of the “third space losses” significantly reduced postoperative complications compared with a standard regimen following traditional guidelines. These results have been confirmed by other prospective randomized trials from more diversified elective intraabdominal procedures. Fluid kinetics studies indicate that the peripheral expandable fluid space during anesthesia and surgery is similar to that in awake volunteers. The peripheral expandable space is in equilibrium with infused fluid, further contradicting the argument of an ECF contraction needing replacement during surgery. Nevertheless, situations with low perfusion might occur in which cell membranes are altered and fluid does indeed shift to the intracellular space. These compartmental shifts might need replacement, but the use of aggressive fluid management in elective surgical procedures is not justified by evidence.

Numerous studies have argued that it should be beneficial to vary the choice of IV fluids in certain situations. Fundamentally, crystalloids (ionic solutions with molecules of <30 kDa) should be better suited for replacing losses from the ECF, whereas colloids with larger molecules should be the ideal replacement for intravascular losses (apart from blood and plasma). Although these fluids show different properties, it has not been possible to show any differences in mortality in large randomized studies. The crystalloid-colloid debate has in recent years focused more on safety aspects such as renal damage in critically ill patients.

Fluid Shifts and Losses During Surgery, and Their Replacement

Because of the porous nature of the endothelium, fluid shifts of protein-poor fluid from the intravascular space to the interstitium are inevitable. Surgical manipulation per se can increase interstitial water volume, and crystalloid infusion can influence its extent. Losses that consistently occur during surgical procedures include urinary output and insensible losses. During trauma surgery and certain other major operations, blood loss constitutes a significant fluid loss. Urinary output and evaporation should affect the extracellular space (vascular system and interstitium) and should cause no net change in colloid osmotic pressure in the vascular system. On the contrary, intravascular loss (bleeding) contains blood components and thereby causes a change in colloid osmotic pressure with crystalloid replacement.

To replace these losses, it is logical to replace extracellular losses with crystalloid. The conventional perception of the distribution of crystalloids is that up to 80% are distributed to the interstitial space. However, fluids are context-sensitive. Advanced kinetic analysis (see later “Body Fluid Dynamics” section) demonstrates a central accumulation of fluid that eventually distributes to the periphery or is eliminated as urine. In simple terms, this means that a crystalloid load initially exerts a substantial volume expansion effect ( Fig. 41.4 ), which is more pronounced during low pressure or bleeding compared with normal states, but the effect is transitory. This, however, has been challenged by those who advocate direct assessment of plasma and blood volume. Colloids, on the other hand, have a more prolonged effect on intravascular volume. To replace blood loss, it is plausible to replace intravascular loss either by giving blood or a colloid. Whole blood would be preferable but this is not usually possible owing to logistical reasons and infectious and incompatibility risks.

Fig. 41.4

Dilution profiles for a crystalloid load. Fluid was given over 10 minutes, reflecting the time course of preload augmentation. Hemoglobin is used as an endogenous tracer to calculate plasma dilution and the data fitted to nonlinear equations. The broken red curve is the kinetic profile of a one-compartment (one volume [ 1V ] space) model, and the continuous blue curve shows the dilution profile of a two-compartment model (two volume [ 2V ] spaces). The dilution is maximal when the infusion is turned off. The effect is transitory and the fluid is almost eliminated from the central space within 3 hours .

Conventional Indices of Resuscitation

Resuscitation of critically ill patients requires accurate assessment of intravascular volume status and the ability to predict the hemodynamic response to a fluid challenge. Indices of hemodynamic response, such as blood pressure, cardiac output, heart rate, and oxygen delivery, do not fully reflect the adequacy of tissue perfusion. Less than 50% of critical care patients given fluid boluses are volume responsive. A reduction in renal perfusion normally results in dilatation of the afferent glomerular arteriole and constriction of the efferent arteriole so that the GFR is kept constant. However, if mean arterial pressure falls below 70 mm Hg (the kidney autoregulatory threshold), renal perfusion pressure and GFR fall, leading to oliguria. However, the kidney is affected by many factors, including cardiac function, osmotic load, intrathoracic pressure, intraabdominal pressure, and chronic renal insufficiency, which make urine output an unreliable predictor of volume status. Other signs of inadequate intravascular volume are peripheral cyanosis, skin mottling, tachycardia, hypotension, and cold extremities. All these signs are nonspecific and unreliable indicators of adequate resuscitation.

Response to Fluid Challenge

An important concept for guiding rational fluid administration is the use of the Frank-Starling curve for cardiac performance ( Fig. 41.5 ). In most clinical circumstances, a basic assumption is that the patient is on the ascending part of the Starling curve and has a submaximal cardiac output. When the subject reaches the flat part of the curve, more fluid administration has little effect on cardiac output/stroke volume and will only increase tissue edema. Although this concept applies to healthy volume-depleted patients, less than half of critically ill patients might respond to a fluid challenge.

Fig. 41.5

Relationship between preload and stroke volume. An incremental increase (A-B) in volume on the x -axis (left ventricular end-diastolic volume, LVEDV ) increases stroke volume/cardiac output on the y -axis ( Responder ). Eventually a deflection point is reached (between C and D), after which the heart will no longer perform increased work with increasing preload ( Nonresponder ).

Static Measurements of Intravascular Volume

Central venous pressure (CVP) is still used as a common parameter to guide fluid therapy ; however, the idea that CVP reflects intravascular volume is a common misconception. CVP was earlier measured in centimeters of water, but now millimeters of mercury is used (1 cm H 2 O = 0.735 mm Hg or 10.2 kPa). CVP is influenced by many factors unrelated to actual fluid balance, such as venous tone, intrathoracic pressure, and left and right ventricular compliance. Consequently, there is poor correlation between CVP and the right ventricular end-diastolic volume, which it is intended to measure. Meta-analyses assessing the ability of CVP to predict fluid responsiveness show low or insignificant predictability in that there is poor association between CVP and circulating blood volume. Chest radiographs and ultrasonography, including vena cava collapsibility index, have limited value in guiding fluid management.

Pulmonary Artery Occlusion Pressure

The pulmonary artery catheter (PAC) is used to measure pulmonary artery occlusion pressure or pulmonary capillary wedge pressure, which is intended to reflect left ventricular preload. However, pulmonary artery occlusion pressure is not a good indicator of preload. The catheter measures pressure and not volume, the relationship is not direct but curvilinear, and clinical benefit is highly doubtful. Left ventricular compliance is dependent on filling of the right ventricle, which means that the PAC has the same limitations as the central venous catheter.

Transesophageal Echocardiography

Widely used in cardiothoracic surgery, transesophageal echocardiography has not proved a reliable predictor of fluid responsiveness in critically ill patients. There are conflicting results for use of the left ventricular end-diastolic area as a good predictor because it is a static measurement. Continuous measurements, such as positive-pressure ventilation, induced changes in vena cava diameter, and aortic flow velocity/stroke volume assessed by echocardiography, all have limitations.

Intrathoracic Blood Volume Index and Global End-Diastolic Volume Index

Transpulmonary thermodilution is a method that uses a cold bolus as a single indicator for determination of the assessment of the largest volume of blood contained in the four heart chambers, called the global end-diastolic volume. It requires the use of a specific thermodilution arterial catheter (pulse contour cardiac output monitoring) that measures temperature changes after the injection of the bolus through a central vein catheter (normally the central vein catheter is placed in the neck, the arterial line in the femoral artery). It does not appear to be a good predictor of fluid responsiveness according to available studies.

Stroke Volume Variation and Pulse Pressure Variation

Intermittent positive-pressure ventilation induces cyclic changes in cardiac loading conditions. Insufflation decreases preload and increases afterload of the right ventricle (RV). The increase in RV afterload is related to inspiratory increases in transpulmonary pressure. The RV preload reduction is due to the decrease in the venous return pressure gradient, which in turn is dependent on the inspiratory increase in pleural pressure. These simultaneous changes in RV preload and afterload lead to a decrease in RV stroke volume, which is at its minimum at the end of inspiration. The inspiratory reduction in RV stroke volume corresponds to a reduction of left ventricle (LV) filling delayed two or three heartbeats because of the filling of the lungs. Thus the LV preload reduction is at its minimum during inspiration. Changes in RV and LV stroke volume are greater when the ventricles work on the steep part of the Starling curve (see Fig. 41.5 ). Thus change in LV stroke volume is an indicator of biventricular preload dependence. According to a large systematic review, pulse pressure variation (PPV) and stroke volume variation (SVV) predict with a high degree of accuracy (receiver operating curve 0.94 and 0.84, respectively) patients who are likely to respond to a fluid challenge. This suggests that currently PPV and SVV are the most accurate, dynamic tools for guidance of fluid management, but they are limited by arrhythmias and the requirement for mechanical ventilation. They are less reliable in critically ill patients receiving ventilatory support. Furthermore, PPV is a direct measurement, whereas SVV is an indirect calculation from pulse contour analysis. PPV can be affected by the tidal volume, with the recommendation that tidal volume should be at least 8 mL/kg. the pleththysmographic variability index is a measure of respiratory-induced variations in the plethysmographic waveform amplitude that, like other dynamic indices, has been shown to predict fluid responsiveness. This is a rather new method provided by noninvasive hemoglobin measurement devices.

Esophageal Doppler Catheter

This technique measures blood flow velocity in the descending aorta by means of a Doppler ultrasound device placed at the tip of a flexible probe. Patients need to be anesthetized or at least sedated to tolerate the probe; once in the esophagus, the catheter is rotated so that the transducer faces the aorta. Because the width of the aorta is either measured or known, the cardiac output can be calculated when the heart rate is known. The flow time corrected for heartbeats is considered an indicator for volume and afterload status. It appears to be a reliable predictor of fluid responsiveness, and has benefits for specific patient groups in reducing morbidity and hospital stay. Performance depends on catheter positioning, but the required training for insertion is short.

Passive Leg Raising Test

Passive leg raising (PLR) represents an endogenous volume challenge that can be used to predict fluid responsiveness. This procedure rapidly returns 150 to 200 mL of blood from peripheral veins in the lower extremities to the central circulation. When ventricular preload is increased, cardiac output is augmented according to the degree of preload reserve.

A PLR test can be used in patients with spontaneous or assisted breathing and irregular cardiac rhythm; therefore it can be a more useful test of fluid responsiveness in critically ill patients. PLR is a simple test in which the legs are elevated by 45 degrees above the bed, but to be reliable it must be accompanied by cardiac output monitoring. Because of the ease of use, simplicity, and high diagnostic accuracy, PLR is the preferred method to assess fluid responsiveness.

Oxygen Delivery and the Microcirculation

The microcirculation consists of myriads of vessels connecting the arteriolar and venous vasculature. It is here that oxygen is delivered and consumed and that metabolism generates waste products. It is important to distinguish between convective oxygen transport, which depends on the availability of red cell oxygen-transporting capacity, and diffusion oxygen transport, which is related to the pressure gradient and inversely related to the distance between the capillary and the cell (specifically mitochondria). Mixed venous oxygenation (SvO 2 ) is a global assessment but requires a PAC. Its surrogate, central venous oxygenation (ScvO 2 ), also requires a central venous line and is a regional index (head and upper body). Although ScvO 2 is slightly higher than SvO 2 in critically ill patients, they usually trend together.

Several options are available to monitor the microcirculation. Brain tissue oxygenation has been measured extensively in experimental animal cardiopulmonary resuscitation models but has limited clinical applicability owing to invasiveness. Among direct methods, in vivo microscopy of the microcirculation is available through small cameras mainly restricted to the sublingual area. Earlier devices such as orthogonal polarized spectral and sidestream dark field imaging video microscopes were hampered by technological limitations, but recent-generation handheld microscopes perform better. These devices make it possible to assess the two determinants of oxygen delivery. Microvascular flow index is a marker for convective oxygen transport and the diffusion distance is reflected by capillary density.

Among indirect methods, tonometry and capnography are based on local accumulation of carbon dioxide as a result of inadequate elimination of cellular waste, and could possibly be a useful method for guiding fluid therapy. However, there is a confounding situation when distributive failure occurs (shunting, such as in sepsis). Another method is the use of near-infrared spectroscopy, a noninvasive optical measurement in which differential absorption of infrared light at two specific wavelengths (680 and 800 nm) by deoxyhemoglobin is used to define hemoglobin saturation level in vessels located under the probe.

Body Fluid Dynamics (Modeling Fluid Therapy)

How do we analyze and quantify the volume and equilibrating process induced by infusing fluid intravascularly? Is it possible to compare effects of different fluid therapies given in various clinical situations? A key problem in studying these phenomena is that infused fluid is added to a highly regulated system that attempts to maintain intravascular, interstitial, and intracellular volume through homeostatic adaptation. We can use dynamic models to predict more accurately the time course of these volume changes. These analyses should permit estimation of peak volume expansion and rates of clearance of infused fluid and covariate analysis of other effects, such as changes in cardiac output or cardiac filling pressures.

The traditional clinical assumption of the distribution of infused crystalloids, one-third intravascular and two-thirds peripheral, is not always accurate owing to large variations between individuals ( Fig. 41.6 ). The response to hypovolemia or hypervolemia depends highly on the starting hydration level, underlying pathology, and general ability to eliminate and distribute fluid via homeostatic mechanisms.

Fig. 41.6

Mass balance computations for fluid infused over 30 minutes. The figures illustrate the distribution of infused Ringer solution in milliliters (dark pink) between plasma (blue line), accumulated urine (purple line), and the peripheral fluid space (yellow line) for two subjects over time in minutes. The subject in the left panel responds quickly to the hypervolemia and has eliminated almost all the infused fluid during the observation time. The subject in the right panel has a less pronounced and slower elimination.

(Reprinted with permission from Rodhe P, Drobin D, Hahn RG, et al. Modelling of peripheral fluid accumulation after a crystalloid bolus in female volunteers—a mathematical study. Computat Math Method Med . 2010;11:341–351.)

Hemoglobin can be used as an endogenous marker to compute intravascular expansion. If the intravascular volume V b (in milliliters) is defined as the distribution volume of hemoglobin (Hb), it is natural to use Hb as an endogenous tracer when analyzing volume expansion; assuming no loss of red cells, as plasma volume expands, Hb concentration decreases. Consider the relationship between the amount of Hb ( X Hb ) and its concentration C Hb (in millimoles per milliliter):

<SPAN role=presentation tabIndex=0 id=MathJax-Element-3-Frame class=MathJax style="POSITION: relative" data-mathml='CHb=XHbVb’>CHb=XHbVbCHb=XHbVb
C H b = X H b V b

If V b is assumed to be a completely closed but expandable space ( Fig. 41.7 ), and fluid is infused by a constant rate R i (in milliliters per minute), the concentration at any time t can be computed as

<SPAN role=presentation tabIndex=0 id=MathJax-Element-4-Frame class=MathJax style="POSITION: relative" data-mathml='CHb(t)=XHbVb+Ri⋅t’>CHb(t)=XHbVb+RitCHb(t)=XHbVb+Ri⋅t
C H b ( t ) = X H b V b + R i ⋅ t

Fig. 41.7

Closed volume approach. Fluid is infused at a rate of R i , which causes an expansion of the intravascular volume V b . This in turn lowers the concentration of the endogenous marker hemoglobin (Hb).

Because X Hb is considered to be constant during infusion, Eqs. (1) and (2) can be combined:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-5-Frame class=MathJax style="POSITION: relative" data-mathml='Vb⋅(CHb(0)CHb(t)−1)=Ri⋅t,’>Vb(CHb(0)CHb(t)1)=Rit,Vb⋅(CHb(0)CHb(t)−1)=Ri⋅t,
V b ⋅ ( C H b ( 0 ) C H b ( t ) − 1 ) = R i ⋅ t ,
rewriting Eq. (3) as
<SPAN role=presentation tabIndex=0 id=MathJax-Element-6-Frame class=MathJax style="POSITION: relative" data-mathml='1Ri⋅(CHb(0)CHb(t)−1)=Y=1Vb⋅t≅θ⋅t+ε’>1Ri(CHb(0)CHb(t)1)=Y=1Vbtθt+ε1Ri⋅(CHb(0)CHb(t)−1)=Y=1Vb⋅t≅θ⋅t+ε
1 R i ⋅ ( C H b ( 0 ) C H b ( t ) − 1 ) = Y = 1 V b ⋅ t ≅ θ ⋅ t + ε

Eq. (4) allows us to compute Y and ε (the residual error) by regression, and to obtain an estimate of theta (1/ V b ). The method is straightforward and only needs Hb data as an input. Fig. 41.8 shows an example of experimental data obtained using the closed volume approach. In the right panel, the Hb infusion estimation overestimates V b by nearly 1 L. Although the predictions, based on gender, height, and weight, can differ considerably between individuals, the estimates based on the closed volume approach should be adjusted by an empirical correction term. This overestimation phenomenon might arise from (1) underestimating the rate of infusion or, more likely, (2) fluid disappearing from the intravascular fluid space during infusion. Therefore the complexity of the fluid distribution model must be increased to understand intravascular volume changes.

Fig. 41.8

An example of data analyzed using the closed volume approach. The analysis is based on data from Fig. 41.7 . Left, Y is estimated by linear regression (solid line) from three consecutive time points ( t = 1.5, 3, 4.5 min; diamonds ) corresponding to three hemoglobin ( Hb ) values for one subject. Distribution volume ( V b ) is then estimated by 1/ y . Right, The distribution volume V b was computed from 14 subjects (circles) receiving 1.5 mL/kg per minute of lactated Ringer solution. Hb values were taken from arterial blood samples. Although there was a 1 : 1 relation between the anthropometric prediction and the Hb infusion estimation, a bias of 770 mL was present. The dotted line represents the line of identity, whereas the solid line represents the linear regression.

Plasma Volume Expansion

The plasma volume, at any time t , can be computed from the hematocrit (HCT) as

<SPAN role=presentation tabIndex=0 id=MathJax-Element-7-Frame class=MathJax style="POSITION: relative" data-mathml='vp(t)=vb(t)⋅(1−HCT(t))’>vp(t)=vb(t)(1HCT(t))vp(t)=vb(t)⋅(1−HCT(t))
v p ( t ) = v b ( t ) ⋅ ( 1 − H C T ( t ) )

If the volume of erythrocytes is considered constant during the observation time T (which is not true if bleeding is present), then the relation

<SPAN role=presentation tabIndex=0 id=MathJax-Element-8-Frame class=MathJax style="POSITION: relative" data-mathml='HCT(t)=HCT(0)⋅Vbvb(t)’>HCT(t)=HCT(0)Vbvb(t)HCT(t)=HCT(0)⋅Vbvb(t)
H C T ( t ) = H C T ( 0 ) ⋅ V b v b ( t )
holds for 0 ≤ t T where V b = v b (0). Furthermore, the amount of Hb is constant and the relation
<SPAN role=presentation tabIndex=0 id=MathJax-Element-9-Frame class=MathJax style="POSITION: relative" data-mathml='vb(t)=CHb(0)⋅VbCHb(t)=CHb(0)⋅VpCHb(t)⋅(1−HCT(0))’>vb(t)=CHb(0)VbCHb(t)=CHb(0)VpCHb(t)(1HCT(0))vb(t)=CHb(0)⋅VbCHb(t)=CHb(0)⋅VpCHb(t)⋅(1−HCT(0))
v b ( t ) = C H b ( 0 ) ⋅ V b C H b ( t ) = C H b ( 0 ) ⋅ V p C H b ( t ) ⋅ ( 1 − H C T ( 0 ) )

Only gold members can continue reading. Log In or Register to continue

Apr 15, 2019 | Posted by in ANESTHESIA | Comments Off on Intravascular Volume Replacement Therapy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access