Clinical Review of Appropriate Uses for Albumin

Albumin was the first natural colloid for clinical use as a volume expander, and it is the standard colloidal agent for comparison with other colloid products. Colloids refer to large molecules that do not pass readily across capillary walls. These compounds exert an oncotic (i.e., they attract fluid) load and are usually administered to restore intravascular volume and improve tissue perfusion.

Albumin may be used to increase intravascular oncotic pressure and thereby expand intravascular volume in patients with hypovolemic shock, severe burn injury, adult respiratory distress syndrome (ARDS), ascites, liver failure, pancreatitis, and in patients undergoing cardiopulmonary bypass.^1,2 Albumin may also be used to treat neonatal hyperbilirubinemia, hypoproteinemia, and nephrotic syndrome.^1,2 However, the relatively high cost of albumin compared to nonprotein colloids (i.e., hetastarch and dextran) and to crystalloid solutions (i.e., lactated Ringer’s solution [LR] and normal saline [NS]) creates a controversy that has divided the opinion of clinicians for years.^1-3

To make an informed decision about the appropriate use of a colloid (e.g., albumin or hetastarch) versus a crystalloid (e.g., normal saline or lactated Ringer’s), clinicians need to appreciate how colloids differ from crystalloids. In patients with intact vascular linings, colloids remain in the intravascular space and cause less edema than do crystalloids. In addition, less is needed to produce a given amount of volume expansion when compared with crystalloids. On the other hand, albumin may cause hypocalcemia and anaphylactic reactions, and it is more costly than crystalloids. This article reviews the mechanism of action, indications for use, dosages, administration techniques, and precautions regarding intravenous albumin in the clinical setting.

Chemistry and Physiology
Colloids are described according to their molecular weight and size. Monodisperse colloidal suspensions have particles of equal weight and size. Polydisperse solutions have a wide range of molecular sizes and weights.

Osmotic pressure is determined by both the kinetic energy of the solute and the amount of solute in solution—not by the solute’s particle size. Therefore, solutes of unequal masses may exert the same osmotic pressure because they possess equal amounts of kinetic energy and are found in similar concentrations in solution. However, the size of the colloidal particles does determine the rapidity with which the solute is cleared from the intravascular space. For colloidal dispersions, particle size is inversely proportional to osmotic pressure and directly proportional to intravascular residence time.⁴

The ideal colloid would produce a sustained increase in plasma volume without posing a risk for anaphylaxis, edema, or infection.4-6 The agent would be readily available, affordable, and convenient to store and deliver. Unfortunately, the ideal colloid does not exist, but human serum albumin (HSA) has a record of proven efficacy and safety when used appropriately.^4-7

Albumin, globulin, and fibrinogen are the principal proteins in plasma.⁴ Under normal physiological conditions, their concentrations in plasma are 4.5 g/dL, 2.5 g/dL, and 0.3 g/dL, respectively.⁴ Albumin is a monodisperse solution with an average molecular weight of 66,300 D to 69,000 D.^4,7Albumin is highly water-soluble and carries a strong negative charge (–17) at pH 6.4 to 7.4.^4,7 This electrical charge explains why albumin is removed rapidly when plasmapheresis is employed to remove substances that are attached to albumin and are detrimental to the body.

Table 1
Potential Albumin Losses in Pathologic States
Fluid	Daily Loss of Albumin
Edema Congestive heart failure Renal disease Cirrhosis	1 g/dL
Lymphedema	>2 g/dL
Ascites	1–2 g/dL
Nephrosis	100–400 mg/kg/d
Source: reference 5

The primary protein in human plasma, albumin comprises 75%–80% of the normal colloid oncotic pressure, and about 50%–60% of the protein content.⁴ One gram of intravascular albumin binds to 18 mL of water by oncotic pressure. Albumin enters the intravascular space via two routes—passing directly from hepatocytes into the sinusoids or traversing the space of Disse (the gaps between hepatocytes and sinusoidal walls) and entering the hepatic lymphatic system and the thoracic ducts. After 2 hours, 90% of albumin remains in the intravascular space. The plasma half-life of albumin is 16 hours, and the reticuloendothelial system plays a major role in the removal of albumin from the body. Daily loss of albumin from the intravascular space is approximately 10%.⁴ Certain pathological conditions increase the daily loss of albumin (TABLE 1).⁵

Albumin is produced in the liver at a rate of 130–200 mg/kg/day.⁵ Catabolic states increase the breakdown of albumin and may lead to stress-induced hypoalbuminemia.⁴ Stress and traumatic injury acutely decrease albumin synthesis, simultaneously increasing the production of acute-phase reactant proteins (e.g., globulin, fibrinogen, and haptoglobulin). If adequate nutrition is present and the liver is functioning normally, the body releases thyroid hormone and cortisol to stimulate RNA production and increase the synthesis of albumin. However, the principal regulator of albumin synthesis is the colloid oncotic pressure (COP) within hepatocytes.^5-7 The serum albumin concentration does not exert a feedback loop to regulate albumin synthesis except to affect oncotic pressure. Therefore, a reduction in serum albumin will not lead to an increase in the synthesis of albumin unless there is a fall in intrahepatocyte oncotic pressures.

Table 2
Distribution of Albumin in the Body
Organ	Amount (g/70-kg man)	Concentration (g/kg organ)
Intravascular
Blood	140	24.0
Extravascular
Muscle	50	2.3
Skin	40	7.7
Liver	2	1.4
Gut	8	5.0
Other tissues	110	3.0
Total extravascular	210
Total body	350
Source: reference 8

Approximately 30%–40% of the body’s albumin is found in the intravascular space (TABLE 2).^5,8 The extravascular space holds approximately 210 g of albumin.⁸ While some extravascular albumin is bound to tissue and metabolized to amino acids, free albumin travels from the extravascular space into the intravascular space via lymphatic drainage. In patients with liver disease and increased portal-systemic pressures, albumin leaves the intravascular space and enters directly into the peritoneal cavity to produce ascites. Loss of albumin from the intravascular space into the peritoneal cavity causes intravascular protein deficiency in patients with liver disease—despite a normal or elevated rate of albumin synthesis.

The normal transcapillary leak of albumin is 5% per hour.⁴ In critically ill patients, loss of capillary wall integrity causes a 60-fold increase in that leak.⁴ The increase affects the equilibrium between plasma albumin and lung lymph albumin. This dynamic equilibrium helps prevent excessive pulmonary edema in response to increased pulmonary vascular resistance, and is vital to the alveolar-capillary exchange of oxygen.⁷ Under normal conditions, the half-life of equilibration between plasma albumin and pulmonary lymphatic albumin is 3 hours; it decreases to 2.5 hours following increased pulmonary vascular pressure. Alterations in pulmonary vascular permeability increase the transcapillary leak rate and reduce the half-life of equilibration to less than 1 hour.⁷ In these situations, HSA should not be used for volume replacement because it carries an increased risk of noncardiogenic pulmonary edema.

An important feature that separates albumin from other colloids or crystalloids is its unique ability to bind reversibly to both cations and anions.⁹ Albumin has three binding sites (an acidic site, a basic site and a neutral site), which play important roles in the transport of lipids and lipid-soluble materials.⁹ Albumin is a carrier molecule for fatty acids, hormones, enzymes, dyes, trace metals, and drugs, and thereby regulates the extracellular concentrations of these substances.^7-9 Because protein-bound substances are rendered inactive, albumin indirectly controls their biological activity.⁸ According to Tullis, the physiologic effects of these binding properties complement the volume-expanding properties of albumin.⁷ Clinicians need to remember that important drug-displacement interactions are influenced by both the fluctuations in serum albumin concentration and by the severity of underlying illness.⁹

Table 3

Human Serum Albumin Products for Clinical Use

Human Serum Albumin (HSA)
25?.5% protein
or
5?.3% protein

Composition: At least 96% of the total protein in the final product is albumin
pH: 6.9 ?0.5
Sodium: 5% HSA 130–160 mEq/L CFR
Expiration: 3 years; stored at 30�C

Plasma Protein Fraction
5?.3% protein

Composition: At least 83% of the total protein in the final product is albumin; no more than 17% globulins; no more than 1% of the globulins are gamma globulin
pH: 7.0 ?0.3
Sodium: 130–160 mEq/L CFR
Expiration: 3 years; stored at 30�C

CFR = Code of Federal Regulations

Commercial Preparations
Albumin for commercial use is obtained from source blood, plasma, or serum of healthy human donors by fractionation according to the Cohn cold ethanol process.⁷ Albumin is pasteurized for 10 hours at 60�C to inactivate human immunodeficiency virus (HIV) and hepatitis viruses. Sodium caprylate and sodium acetyltryptophanate are added to prevent denaturation during pasteurization. TABLE 3 compares the three commercially available forms of albumin (5% albumin, 25% albumin, and plasma protein fraction).⁵ If stored correctly at 30�C, these products have a shelf-life of 3 years. The 25% solution of albumin is hypertonic with an osmolarity of 1,500 mOsm/L, and it produces a four- to five-fold expansion of intravascular volume. The 5% solution is isotonic with blood. Human serum albumin contains at least 96% albumin, whereas plasma protein fraction (PPF) contains at least 83% albumin. Neither HSA nor PPF contains clotting factors.

Human serum albumin is safe to administer and rarely causes adverse effects. The incidence of urticaria, fever, chills, or hypotension ranges from 0.47%–1.53%.7 Overall incidence of allergic reactions to HSA is 0.011% and of contamination with hepatitis is negligible.⁴ Because PPF contains greater quantities of kinins or prekallikrein, it carries a greater risk of hypotension than does HSA. Premature infants and patients receiving long-term parenteral nutrition may be at increased risk of aluminum toxicity since albumin contains trace amounts of aluminum.¹⁰
As previously stated, hypocalcemia may arise when albumin therapy is used for fluid resuscitation. Albumin avidly binds calcium and reduces the fraction of ionized calcium. The reduction in ionized calcium produces myocardial depression.¹¹ When compared with crystalloid therapy in trauma patients, albumin-treated patients required more intravascular fluid replacement, produced less urine output, and had decreased renal function.¹² Gore and colleagues reported similar paradoxical decreases in urine output following fluid resuscitation with albumin in burn injury patients.¹³ Studies have examined the effect of albumin on glomerular filtration rate (GFR).^13,14 The administration of 25% albumin to patients with burn injury resulted in a 40% increase in plasma volume and a paradoxical decrease in GFR and urine output.¹³ The mechanism of albumin-induced reduction in GFR is unknown. Studies show that albumin increases the oncotic pressure within the peritubular vessels, causing a decrease in sodium and water excretion.¹⁴

Albumin’s Role in the Body
The albumin portion of human blood serves three primary physiologic roles: 1) maintenance of plasma colloid osmotic pressure, 2) transport and sequestration of bilirubin, and 3) transport of fatty acids and other intermediate metabolites such as hormones and enzymes.⁸ Because albumin accounts for approximately 80% of the oncotic pressure of plasma, a 50% reduction in serum albumin concentration produces a 66% decrease in colloid oncotic pressure.⁵ In critically ill patients, risk of death is inversely related to serum albumin concentration.¹ Goldwaser and Feldman estimate that for each 2.5 g/dL decrease in serum albumin concentration, there is a 24%–56% increase in the risk of death.⁶ This estimate was made after adjusting for other covariates (e.g., renal function, serum transaminase, lactic acidosis), and it suggests that albumin may have a direct cytoprotective effect.^1,6

Because albumin plays a significant role in regulating blood volume, its primary place in therapy is volume expansion in hypovolemic shock, hemorrhage, sepsis, and cardiopulmonary resuscitation.

Clinical Use
The 5% and 25% concentrations of albumin are indicated for treatment of hypovolemia with or without shock. Albumin restores intravascular volume and maintains cardiac output and colloid osmotic pressure. A 25 g dose provides the osmotic equivalent of 2 units (500 mL) of fresh-frozen plasma. A 500 mL solution of 5% albumin (25 g of albumin) increases plasma volume by 450–500 mL. When plasma volume is dramatically reduced, albumin can restore the losses quickly in most cases. Administering 100 mL of a 25% solution of albumin (25 g) draws 350 mL of interstitial fluid, to yield an intravascular volume of approximately 2 pints of whole blood (450 mL) over 30–60 minutes. If blood loss is severe, a transfusion of whole blood or packed red cells may be needed to restore hematocrit and improve oxygen delivery. Advantages of albumin over whole blood include the following: no need for type- and cross-matching, negligible risk of contracting blood-borne diseases, and absence of sensitization with repeated administration.¹⁵

A colloid oncotic pressure of Ž20 mmHg, a serum albumin of Ž2.5 g%, or a total serum protein of Ž5 g% indicates adequate plasma oncotic activity.⁵ The administration of crystalloids reduces COP. In contrast, the administration of colloids increases COP. Studies have evaluated the relationship between COP and postoperative recovery in patients requiring intraoperative fluid replacement.^16,17 Albumin was administered to achieve a target COP of 24 mmHg (Group 1) or 29 mmHg (Group 2).¹⁷ Group 1 received a mean of 51.7 ?81.1 g of albumin and Group 2 received 95.127 ?123.1 g of albumin (p = 0.0018). There was no significant difference between groups with regard to length of stay, postoperative complications, or duration of ventilatory support.¹⁷ Two other studies evaluated COP and pulmonary capillary wedge pressure (PCWP) as a means of assessing albumin requirements in critically ill patients with hypovolemia.^18,19 Although albumin improves COP in patients with hypovolemia, measuring COP and PCWP to gauge albumin requirement is of limited clinical value. Fluid may move from the intravascular to the extravascular space without a clinically significant change in either COP or PCWP.^18,19

Comparison of Colloids and Crystalloids
Several colloids and crystalloids are available for expanding intravascular volume in various settings (TABLE 4).²⁰ Crystalloids consist of electrolytes (e.g., sodium, chloride, and potassium) in water solutions.²⁰ The most frequently used crystalloids are normal saline and lactated Ringer’s. These crystalloids are isotonic and rapidly distribute into extracellular water. There is no clinically relevant difference in safety and efficacy between LR and NS, but LR should be avoided in patients with liver failure because it impairs metabolism of lactate to bicarbonate and may result in lactate accumulation. Advantages of crystalloids over colloids include ease of peripheral administration because of their isotonicity, lower risk of fluid overload, negligible effect on COP, freedom from anaphylactoid reactions, and affordability. The main disadvantage of crystalloids is that large volumes are needed to re-establish intravascular volume. Advantages of colloids are their long dwell time in the vasculature and the lower volume requirement necessary to produce a given amount of intravascular volume expansion. Albumin, dextran, plasma protein fractions, and hetastarch should be administered through a central venous line to limit the risk of phlebitis and to allow for dilution of the viscous colloidal solutions.

Table 4
Comparison of Blood Volume Expanders
	HSA 5%	HSA 25%	Dextran 40	Frozen Plasma	Hetastarch	PPF	Lactated Ringer’s	Saline
Primary use	Volume expansion, hypotension	Volume expansion, hypotension	Volume expansion, hypotension	Coagulopathy	Volume expansion, hypotension	Volume expansion	Volume expansion, hypotension	Volume expansion, hypotension
Other use	Protein deficiency, kernicterus	Protein deficiency, kernicterus	Anticoagulant	TTP therapy	Anticoagulant	Coagulopathy	Routine hydration, electrolyte	Routine hydration, hypercalcemia supplement
Contraindication	Allergy	Allergy	Allergy, bleeding	Anti-IgA	Allergy, bleeding	Allergy	Hypervolemia, liver failure	Hypervolemia
Side effects	Hypervolemia, hypoCa++	Hypervolemia, hypoCa++	Coagulopathy, hypervolemia, bleeding, allergic rxn	None	Coagulopathy, hypervolemia	Allergy, bradycardia, hypotension	Hypervolemia, metabolic acidosis	Hypervolemia, hypernatremia
Anaphylactoid rxn	0.019%	0.019%	3%	0.019%	0.085%	0.019%	0%	0%
Patient profile	Surgery, trauma, burn, peds,* ascites sepsis, liver failure	Surgery, trauma, burn, peds,* ascites sepsis, liver failure	Surgery, trauma, sepsis, liver failure	Cancer, coagulopathy, liver failure, burn	Surgery, trauma, sepsis, burn	Coagulopathy, surgery, trauma,	Surgery, trauma, sepsis, burn, diabetes, CPR	Surgery, trauma, sepsis,burn, diabetes, CPR
Sodium (mEq/L)	130–160	130–160	154	130–160	154	130–160	130	154
Chloride (mEq/L)	130–160	130–160	154	130–160	154	130–160	109	154
Other					Hydroxyethyl starch MW 450,000 D		Lactate 28 mEq/L, K⁺ 4 mEq/L, Ca⁺⁺ 3 mEq/L
COP (mmHg)	20	100	30	20	30	20	0	0
Osmolality (m0sm)	300	1,500	300	300	310	290	273	308
Plasma expansion (per person)	240–500 mL (per 250 mL)	250 mL (per 50 mL)	750 mL (per 500 mL)	250–500 mL (per 250 mL)	750 mL (per 500 mL)	250–500 mL (per 250 mL)	250 mL (per 1 L)	250 mL (per 1 L)
Half-life	5–6 hours	5–6 hours	4–6 hours	5–6 hours	6–8 hours	5–6 hours	Minutes	Minutes
Elimination	Hepatic	Hepatic	Renal	Hepatic	Renal	Hepatic	Hepatorenal	Renal
Source	Human	Human	Synthetic	Human	Synthetic/botanical	Human	Crystalloid	Crystalloid
Adapted from reference 20 *peds = pediatric patients

The Cochrane Injuries Group Albumin Reviewers recently published a meta-analysis of 30 randomized, controlled trials that compared the administration of albumin or plasma protein fraction with crystalloids in 1,419 critically ill patients with hypovolemia, burns, or hypoalbuminemia.¹ The outcomes measure of the meta-analysis was mortality from all causes at the end of follow-up for each trial. Among patients with hypovolemia, albumin increased the relative risk of death by 1.46 (95% CI, 0.97 to 2.22).¹ For patients with burn injury or hypoproteinemia, albumin increased the relative risk of death by 2.40 (95% CI, 1.11 to 5.19), and by 1.68 (95% CI, 1.26 to 2.23), respectively.¹ Pooled data showed that the administration of albumin increased mortality by 6% (95% CI, 3 to 9). Cochrane and coauthors concluded that albumin has not been shown to improve outcome in critically ill patients, and they cautioned against widespread use of albumin in this patient population.¹ A critique of the Cochrane retrospective meta-analysis stated that “the use of albumin, the purity of albumin, and the methods of monitoring fluid resuscitation have improved since the trials reviewed by this meta-anaylsis were conducted.”²¹ According to this reasoning, there is no current scientific basis for discarding the use of albumin for fluid resuscitation in appropriately selected and appropriately monitored patients.²²

Recently, Shoemaker used continuous monitoring of hemodynamics and oxymetrics with electrical bioimpedance to show that albumin was significantly better than crystalloids for improving blood pressure, cardiac output, and tissue perfusion in critically ill patients.²² Further, two recent studies show that IV infusions of albumin could be used safely and effectively in patients with cirrhosis and ascites.^23,24 In patients with ascites, IV albumin infusions compared favorably with surgical intervention.²⁴

Although the controversy regarding the agent of choice for fluid resuscitation remains unsettled, there are distinct clinical situations in which albumin or a nonprotein colloid would be preferred over a crystalloid solution.^2,20-24 Patients with fluid accumulation in the third space (e.g., ascites) and intravascular volume depletion benefit from the administration of albumin or other colloids because these plasma volume expanders augment intravascular volume, draw fluid out of the third space and shift fluid back into the vasculature. However, overzealous intravascular volume replacement carries the risk of congestive heart failure and hypervolemia; therefore, therapy must be administered with care. In patients with impaired capillary walls, colloids should be used with great caution, if at all, to prevent capillary leak, accumulation of fluid in the third space, and pulmonary edema.²⁰

Standard of care dictates that, in most cases, crystalloids be the agents of choice for fluid resuscitation.^1,2,20,25 This standard is based on the ease of administration, safety, and economic benefit of this approach.^1,2,20,25 Colloid administration should be reserved for clinical situations in which fluid resuscitation with crystalloids has failed to reduce the intravascular volume deficit.² Because colloids produce a dilutional effect on hemoglobin and on clotting factors, clinicians need to monitor the appropriate parameters to safeguard against iatrogenic complications.^20,21 Electrical bioimpedance devices that continuously measure cardiopulmonary hemodynamics and oxymetrics help optimize fluid resuscitation with colloids in critically ill patients.²¹

Guidelines for Use
Albumin is a volume expander and an antihyperbilirubinemic agent. In 1977, the National Institutes of Health (NIH) published guidelines for the therapeutic use of albumin.²⁵ According to the guidelines, only a few disorders justify the regular use of albumin; many conditions justify its occasional use; and in certain situations albumin therapy is unjustified (TABLE 5).²⁵

Table 5
Indications for Human Albumin Therapy
Appropriate Use	Occasional Use	Unjustified Use
Hypovolemia Severe burns CABG surgery ARDS	Acute liver failure Hypoproteinemia Ascites RBC resuspension Acute nephrosis Renal dialysis Detoxification	Chronic cirrhosis Chronic nephrosis Undernutrition
Source: reference 25

More recently, Vermeulen and colleagues conducted a review of the biomedical literature published between 1977 and 1993 on the use of albumin, nonprotein colloids, and crystalloid solutions.² Reviewers used the Delphi method to formulate consensus opinion and practice guidelines for the clinical use of the agents.² Thirty-one allied health professionals, who represented 26 of the 65 University Hospital Consortium (UHC) member institutions, participated in the consensus exercise.² Participants were chosen on the basis of their research expertise and clinical acumen in the use of albumin, nonprotein colloids, and crystalloid solutions.² Of the 31 initial participants, 24 successfully completed the exercise.² TABLE 6 lists the resulting consensus guidelines.²

Dosage and Administration
Albumin is administered via intravenous infusion. Dosage is based on patient needs and response to therapy. Once euvolemia is re-established, the albumin should be stopped. Indicators of euvolemia include normalization and stabilization of blood pressure, return of a palpable pulse, stabilization of heart rhythm and rate, and restoration of normal urinary output. No more than 250 g (5 L of 5% albumin or 1 L of 25% albumin) should be administered over a 48-hour period. Additional fluid requirements should be met by crystalloid replacement or whole blood supplementation. During hypovolemic shock, the rate of infusion of albumin should be between 5 and 10 mL/min; for patients with near normal plasma volumes the rate should not exceed 2–4 mL/min.

Table 6
University Hospital Consortium Guidelines for the Use of Albumin, Nonprotein Colloids, and Crystalloid Solutions
Indication	Guideline
Hemorrhagic shock	Crystalloids should be considered the initial resuscitation fluid of choice; colloids are appropri-ate for resuscitation in conjunction with crystalloids when blood products are not immediately available; on the basis of cost-effectiveness considerations,* nonprotein colloids are favored over albumin, except in the following cases: If sodium restriction is required, the use of 25% albumin diluted to 5% with 5% dextrose solution is recommended; If nonprotein colloids are contraindicated, use of 5% albumin solution is recommended;^† Crystalloid and colloid solutions should not be considered substitutes for blood or blood components when oxygen-carrying capacity is reduced and/or when replenishment of clotting factors or platelets is required; patients who experience symptoms of shock while undergoing hemodialysis are included in this guideline and should receive crystalloid solutions as the resuscitation fluid of choice
Nonhemorrhagic (maldistributive) shock	Crystalloids should be considered first-line therapy for nonhemorrhagic shock; the effectiveness of colloid solutions in the treatment of sepsis has not been demonstrated in clinical trials; however, in the presence of capillary leak with pulmonary and/or peripheral edema, or following the administration of at least 2 L of crystalloid solution without effect, nonprotein colloid may be used; if nonprotein colloids are contraindicated, albumin may be used
Hepatic resection	Crystalloid use to maintain effective circulating volume following major hepatic resection (>40% resected) is recommended; the use of nonprotein colloid solutions and albumin is also appropriate, depending on the function of the residual liver and hemodynamic status; if crystalloids are not used, nonprotein colloids are recommended as the most cost-effective alternative
Thermal injury	Crystalloid solutions should be used for initial fluid resuscitation (within the first 24 hours); colloids should be administered in conjunction with crystalloids if all of the following are true: burns cover more than 50% of the patient’s body surface; at least 24 hours have passed since the burn occurred; crystalloid therapy has failed to correct hypovolemia; based on cost-effectiveness considerations, nonprotein colloids are recommended; if nonprotein colloids are contraindicated, albumin may be used
Cerebral ischemia	Colloid solutions are of no demonstrated value and should not be used in the treatment of ischemic stroke or subarachnoid hemorrhage; their use for these indications should be discouraged, except in patients with hematocrit levels lower than 40% on admission; patients with elevated hematocrit levels on admission should receive crystalloid solutions to increase intravascular volume, creating a state of hypervolemia and hemodilution (hematocrit levels of approximately 30% to maximize cerebral perfusion); additional interventions (e.g., blood removal) may be needed in such cases; colloid solutions (both nonprotein and albumin) should be discouraged on the basis of cost-effectiveness considerations
Nutritional intervention	Albumin should not be used as a supplemental source of protein calories in patients requiring nutritional intervention; however, patients with diarrhea associated with enteral feeding intolerance may benefit from the administration of albumin if all of the following conditions are met: Significant diarrhea (>2 L/d) occurs; Serum albumin is less than 20 g/L (2.0 g/dL); Continued diarrhea occurs despite trial of short-chain peptide and elemental formulas; Other causes of diarrhea have been considered and ruled out
Cardiac surgery	Crystalloids should be the fluid of choice as the priming solution for cardiopulmonary bypass pumps; the use of nonprotein colloids in addition to crystalloids may be preferable in cases in which it is extremely important to avoid pulmonary interstitial fluid accumulation; for postoperative volume expansion, crystalloids should be considered the first-line solution, followed by nonprotein colloids, and finally albumin; nonprotein colloids may be beneficial if reduction in systemic edema is required
Hyperbilirubinemia newborn	Albumin should not be administered in conjunction with phototherapy; albumin should not be used prior to exchange transfusion; albumin has been used with mixed results as an adjuvant to exchange transfusions and should be administered only with concurrent transfusion of blood; crystalloids and nonprotein colloids do not have bilirubin-binding properties and should not be considered as alternatives to albumin
Cirrhosis paracentesis	Albumin, administered alone or in conjunction with diet modification and diuretics, should be avoided for the treatment of cirrhosis with ascites removal of less than 4 L; crystalloids should be considered the solution of choice to prevent complications associated with large-volume paracentesis, such as reduced effective plasma volume, renal dysfunction, and so on; nonprotein colloids and albumin should be considered second-line agents for the prevention of complications following the removal of 4 L or more of ascitic fluid
Nephrotic syndrome	Short-term albumin use, in conjunction with diuretic therapy, is appropriate for patients with acute, severe peripheral or pulmonary edema
Organ transplantation	Albumin and/or nonprotein colloid administration have not been demonstrated conclusively to be effective during and/or after renal transplantation surgery; albumin may be useful for postoperative liver transplant patients in the control of ascites and peripheral edema if all of the following conditions are met: Serum albumin is less than 25 g/L (2.5 g/dL); Pulmonary capillary wedge pressure is less than 12 mmHg; Hematocrit is more than 30%; In these cases, albumin may also be used to replace ascitic fluid lost through drainage catheters following liver transplantation; the use of albumin in liver transplantation is not well documented in the biomedical literature
Plasmapheresis	Albumin, in conjunction with large-volume plasma exchange, is appropriate; large-volume plasma exchange is defined as more than 20 mL/kg in one session, or more than 20 mL/kg per week in repeated sessions; crystalloid solutions and albumin/crystalloid combinations should be considered as cost-effective alternatives for small-volume exchanges
Indications with limited or inconclusive published supportive evidence, considered appropriate based on the results of the consensus exercise	Granulocytapheresis: nonprotein colloid solution is appropriate as a sedimenting agent for donation of granulocytes and for acute cytareduction in chronic myelogenous leukemia (chronic granulocytic leukemia); Stem cell cryopreservation: nonprotein colloid solution is appropriate as part of a cryopreservation solution for frozen storage of hematopoietic stem cells; pretreatment of Dacron aortic grafts: albumin is appropriate to make grafts impervious to blood before insertion; red blood separation for major blood type incompatible bone marrow transplan tation: nonprotein colloids are appropriate; severe, necrotizing pancreatitis: albumin is appropriate
Indications with inconclusive published supportive evidence, considered inappropriate based on the results of the consensus exercise	Severe hypoalbuminemia; impending hepatorenal syndrome; increasing drug efficacy; limited or uncomplicated pancreatitis
* Therapeutic equivalence between products has been identified in several guidelines. In those instances, products are recommended based on economic considerations. Thus, nonprotein colloids (which are currently less expensive than albumin) are preferred over albumin. Changes in the relative cost of these products (e.g., albumin becoming less costly than nonprotein colloids) should be reflected through modifications to the guidelines. † Relative contraindications to the use of nonprotein colloids include, but may not be limited to the following: (1) previous hypersensitivity to the components of the solution, (2) underlying bleeding disorders, (3) risk of serious intracranial hemorrhage, and (4) renal failure with either oliguria or anuria Reprinted with permission: Arch Int Med, Feb. 27, 1995, Vol. 155, p. 377 ?1995 American Medical Association

Albumin should be avoided in patients with severe anemia, congestive heart failure, chronic nephritis, or allergy to albumin. Albumin produces no beneficial effect in patients with nephrotic edema and is not indicated for the improvement of blood pressure in euvolemic patients. Albumin is physically incompatible with verapamil, hydroalcoholic solutions, and fat emulsions.²¹

The FDA is aware of four cases of hemolysis since 1994 that occurred during or following plasmapheresis when human albumin 25% was diluted to a 5% concentration using sterile water for injection. In response, the FDA recommends that manufacturers of human albumin 25% revise product package inserts to include a warning about the risk of potentially fatal hemolysis and actue renal failure when sterile water for injection is used as a diluent. The FDA also recommends that inserts include information on acceptible diluents, such as 0.9% saline or 5% dextrose in water.

Clinical Monitoring of Therapy
Intravascular volume is monitored to ensure adequate hydration. Hemoglobin and hematocrit are monitored to safeguard against dilution hypervolemia. To prevent iatrogenic complications from fluid resuscitation, clinical laboratory monitoring should assess concentrations of sodium, chloride, and calcium. Physical exam should assess vital signs to ensure adequate blood pressure, normal heart rate and respiratory rate, and should uncover signs or symptoms of altered sensorium, jugular venous distention, dyspnea, pulmonary edema, orthostasis, hypertension, peripheral edema, oliguria, and dehydration. Special care must be taken when administering albumin to elderly patients because they are more susceptible to fluid overload and pulmonary congestion. When a patient’s intravascular volume status is unclear, a pulmonary artery catheter or a central venous line should be used to monitor fluid replacement and avoid hypervolemia.

Conclusion
While the colloid/crystalloid controversy remains unresolved, clinical uses for albumin are justified in some settings. The decision to use albumin must be made with careful consideration of the risks of capillary leak, interstitial edema, and dilutional effects on hemoglobin and clotting factors. In appropriately selected and carefully monitored patients, albumin offers advantages over crystalloids.

References
1. Cochrane injuries group albumin reviewers. Human albumin administration in critically ill patients: systematic review of randomized controlled trials. Br Med J. 1998;317:235-240.
2. Vermeulen LC, Ratko TA, Erstad BL, et al. The University Hospital Consortium guidelines for the use of albumin, nonprotein colloids, and crystalloid solutions. Arch Intern Med. 1995;155:373-379.
3. Copelin C. Practical points in the use of albumin for hypovolemia. J Peranesth Nurs. 1998;13:118-120.
4. Roberts JS, Bratton SL. Colloid volume expanders: Problems, pitfalls, and possibilities. Drugs. 1998;55:621-630.
5. Rainey TG, Read CA. Pharmacology of colloids and crystalloids. In: Chernow B (ed). The Pharmacologic Approach to the Critically Ill Patient Williams and Wilkins, Baltimore MD. 1994:272-290.
6. Goldwaser P, Feldman J. Association of serum albumin and mortality risk. J Clin Epidemiol. 1997;50:693-703.
7. Tullis JL. Albumin. Background and use. JAMA. 1977;237:355-360.
8. Peters T. Serum Albumin
9. Emerson TE. Unique features of albumin: A brief review. Crit Care Med. 1989;17:690-693.
10. Klein GL. The aluminum content of parenteral solutions: current status. Nutr Rev. 1991;49:74-79.
11. Kovalik SG, Ledgerwood AM, Lucas CE, et al. The cardiac effect of altered calcium hemostasis after albumin resuscitation. J Trauma. 1981;21:275-279.
12. Ledgerwood AM, Lucas CE. Postresuscitation hypertension, etiology, morbidity, and treatment. Arch Surg 1974;108:531-538.
13. Gore DC, Dalton JM, Gehr TW. Colloid infusions reduce glomerular filtration in resuscitated burn patients. J Trauma. 1996;3:356-360.
14. Lucas CE. Renal considerations in the injured patient. Surg Clin North Am. 1982;62:133-148.
15. Ring J, Messmer K. Incidence and severity of anaphylactoid reactions to colloid volume substitutes. Lancet. 1977;1:466-469.
16. Ersted BL, Gales BJ, Rappaport WD. The use of albumin in clinical practice. Arch Intern Med. 1991;151:901-911.
17. Grundmann R, Heistermann S. Postoperative albumin infusion therapy based on colloid osmotic pressure. Arch Surg. 1985;120:911-915.
18. Grootendorst AF, van Wilgenburg MGM, de Laat PJH, et al. Effect of large volume replacement with balanced electrolyte solutions on extravascular lung water in surgical patients with sepsis syndrome. Intensive Care Med. 1989;15:505-510.
19. Skillman JJ, Restall S, Salzman EW. Randomized trial of albumin vs. electrolyte solutions during abdominal aortic surgery. Surgery. 1975;78:291-303.
20. Sypniewski E. Hypovolemic and cardiogenic shock. In: DePiro JT, et al. (eds). Pharmacotherapy: A Pathophysiologic Approach Appleton Lange, Stamford, CT; 1997:509-541.
21. Offringa M. Consider validity, clinical relevance, and applicability of albumin for critically ill patients. Lancet. 1998;317:343.
22. Shoemaker WC. Circulatory effects of whole blood, packed red cells, albumin, starch, and crystalloids in resuscitation of shock and acute illness. Vox Sanguinis. 1998;74(suppl 2):69-74.
23. Stumpf JL, Palmieri JA, Nostrant TT, Ryan ML. The use of albumin for large volume paracentesis of cirrhotic ascites. Hosp Pharm. 1994;29:461-468.
24. Gines P, Arroyo V, Vargas V, et al. Paracentesis with intravenous infusion of albumin with peritoneovenous shunting in cirrhosis with refractory ascites. N Engl J Med. 1991;325:829-835.
25. Tullis JL. Albumin. Guidelines for Clinical Use. JAMA. 1977;237:460-463.