Document 178306

Printer: Yet to come
October 25, 2008
The ability to administer blood products is a critically important therapeutic modality in the care of patients with acute and
chronic problems. When carried out with a thorough, up-todate understanding of indications, risks, and benefits, blood
transfusion is exceedingly safe and effective. Physicians encounter a large spectrum of medical and surgical conditions
requiring transfusion therapy, including acute blood loss, catastrophic illness in the critical care setting, diseases associated
with chronic anemia, and a variety of congenital and acquired
bleeding disorders. The modern-day care of the critically ill requires a thorough knowledge of the pathophysiology of blood
loss and anemia, as well as an understanding of normal hemostatic mechanisms and the sometimes complex disorders of coagulation encountered in these populations.
In this chapter, the basic concepts of acute blood loss are discussed, and the indications for and use of blood components,
potential risks of blood products, and alternatives to blood
transfusion are reviewed. Because blood products are a limited
resource with potential serious adverse side effects, knowledge
of appropriate indications, potential risks, and available alternatives should allow clinicians to exercise judgment in using
this important resource. Based on the accumulating evidence,
special emphasis will be placed on minimizing transfusion in
the critical care setting.
The ability to transfuse blood safely and successfully is a relatively recent medical advance. Early historical references to
the use of bloodletting and phlebotomy were common, and
were applied to many diseases and disorders. It is possible that
salutary effects were realized in some situations, such as congestive heart failure, but the vast majority of these applications
were based on medical ignorance and likely resulted in harm
to unsuspecting patients.
In February 1666 in Oxford, England, Richard Lower
demonstrated what is thought to be the first known successful transfusion on an animal. The technical details were published in the Philosophical Transactions of the Royal Society
within a year of the experiment. Another Englishman, Francis
Potter, may have preceded him with transfusions to animals,
and possibly to humans, some years prior (1,2). Jean-Baptiste
Denis is credited with the first transfusion to a human in 1667
performed in France. Denis gave 3 pints of sheep blood to a
patient without ill effects. A subsequent attempt to give blood
to the same man “to mollify his fiery nature” led to the patient’s death shortly after the transfusion. A lawsuit resulted,
and Denis went to trial but was ultimately exonerated. The
Paris medical faculty then forbade blood transfusion, which
led to bans on transfusion throughout Europe that lasted until modern times. An 1825 medical journal credited Dr. Philip
Syng Physick of Philadelphia with blood transfusion to a patient, possibly the first record of successful transfusion of human blood (3). In 1828 in England, Blundell administered a
small amount of human blood to a patient with postpartum
hemorrhage, apparently small aliquots from himself, the husband, and another man (4). The patient reportedly felt better,
but it is likely that the small-volume transfusion had little impact on her outcome. In fact, the patient was fortunate not to
have suffered a serious transfusion reaction.
The routine, safe administration of blood products required
several important scientific advances. The discovery of the A,
B, and O blood types by Karl Landsteiner in 1900 and the AB
blood type by Alfred Decastello and Adriano Sturli in 1902 began the era of modern blood transfusion. The first blood bank
was established in 1932 in a Leningrad hospital. The first blood
bank in the United States was established by Bernard Fantus
in 1937 at Cook County Hospital in Chicago. By the 1940s,
techniques of cross-matching, anticoagulation, and storage of
blood, and the establishment of blood banks made routine
blood transfusion a reality. The introduction of plastic storage containers in 1950 and the introduction of refrigerated
centrifugation instruments in 1953 made component therapy
possible (5).
Approximately 14 million units of red blood cells (RBCs)
(packed RBCs and whole blood), 9,875,000 units of platelets,
and 4 million units of plasma are transfused annually in the
United States (6,7). This represents an 11.8% increase since
1999 and a 56% increase since 1980 (8). The use of other
components, especially platelets, has also increased. Because
only about 5% of eligible donors ever donate blood, future increases may exacerbate shortages, especially as the U.S. population ages. Transfusion rates in the United States for 2001 have
been estimated at 48.75 units of red cell transfusion per 1,000
population as compared to 44.93 units of red cell transfusion
per 1,000 population in England, 28 units of red cell transfusion per 1,000 population in Australia, and in 54.8 units of red
cell transfusion per 1,000 population in Denmark (9).
As anemia in critical care illness is common, 25% to 37%
of patients receive at least one blood transfusion during their
intensive care unit (ICU) stay (10–12). In one study (10), 85%
Printer: Yet to come
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
of patients with an ICU length of stay greater than 1 week received at least one blood transfusion. Notably, blood transfusion was not associated with acute blood loss in over two thirds
of these cases. Phlebotomy and decreased production of blood
cells have been implicated as significant contributors to anemia
in the ICU. Since many studies have estimated daily blood loss
from phlebotomy to be at least 40 mL/day (10,11,13), critical care practitioners should carefully consider the need for
frequent blood draws in the ICU.
Collection and Preparation
of Blood Products
Modern-day blood banks have adopted component therapy
to optimize management of the blood supply. Blood is collected from donors and is then separated into its individual
components—packed RBCs, plasma, platelets, and proteins—
to maximize the benefits of each donated unit while minimizing the risk to recipients of blood products. Blood is collected
from donors into plastic bags containing a citrate solution that
binds calcium, thus preventing coagulation. These solutions include citrate phosphate dextrose (CPD), citrate phosphate double dextrose (CP2D), and citrate phosphate dextrose adenine
(CPDA-1). Additional solutions are now available that extend
the shelf life of packed RBCs, and contain dextrose, adenine,
sodium chloride, and either phosphate (AS-3) or mannitol (AS1 and AS-5). After collection, each unit is gently centrifuged to
pack the RBCs, leaving approximately 70% of the platelets
suspended in plasma; the platelet-rich plasma is removed and
centrifuged again to sediment the platelets. All but a small
amount of the resulting supernatant plasma is removed and
rapidly frozen. The platelets are then resuspended, yielding a
platelet concentrate. When the frozen plasma is stored at less
than 18◦ C, it is referred to as fresh frozen plasma. If the frozen
plasma is allowed to thaw at 4◦ C, the precipitate that remains
can be collected to yield cryoprecipitate. Albumin and other
proteins can then be extracted from the remaining plasma.
Another option for the collection of blood leukocytes,
platelets, or plasma is through automated cell separators
(apheresis). Blood is withdrawn from a donor and separated by
centrifuge, and the desired component is removed. The remaining blood is returned to the donor. Using this technique, many
units of leukocytes or platelets can be quickly removed, allowing blood banks to offer products such as single-donor platelet
packs. The administration of a single-donor unit of platelets is
advantageous since it exposes the recipient to only one person’s
antigens, whereas an equivalent dose of pooled platelet transfusion (“six pack” or “ten pack”) exposes the patient to six or
ten sets of antigens, respectively, making subsequent platelet
transfusion less effective since antibodies are formed against
the wide array of foreign antigens. In addition, bacterial contamination is less likely with single-donor apheresis platelets.
Storage Lesion
Storage and refrigeration create progressive changes in packed
RBCs, known as the storage lesion (14). These changes include
an increase in the concentration of potassium, phosphate, and
ammonia; decrease in pH; altered affinity of hemoglobin for
oxygen; changes in RBC deformability; hemolysis; develop-
ment of microaggregates; release of vasoactive substances; and
denaturation of proteins. In addition, the life span of RBCs
becomes shorter the longer cells are stored. This is associated
with a decrease in both intracellular 2,3-diphosphoglycerate
(2,3-DPG) and adenosine triphosphate (ATP). The transfusion
of large volumes of cold blood contributes to the development
of hypothermia, one of the most clinically significant effects of
storage on subsequent transfusion. With the exception of hypothermia, it is important to realize that many of these changes
may be reversed shortly after transfusion, and may, in some
cases, cause metabolic effects that are different from those predicted based on the above ex vivo observations. It is therefore
critical not to empirically treat the theoretically anticipated effects of blood transfusions using “cookbook” approaches (such
as giving one ampule of bicarbonate and one ampule of calcium
with every “x” units of blood). Some of these treatments may,
in fact, be harmful for the patient in hemorrhagic shock.
Administration of Blood Products
Transfusion based on sound physiologic principles and an understanding of relative risks and benefits should give maximal
benefit to the patient, with efficient use of a valuable and finite
resource. Utilizing data from recent studies, it is increasingly
possible to base transfusion practice on scientific grounds. The
most prominent example is the progressive abandonment of
the “10/30” transfusion “trigger” for red cell transfusion in
favor of lower transfusion triggers and, even more appropriately, transfusion practice based on patient physiology (10,15).
The 10/30 transfusion trigger for red cell transfusion likely resulted from a recommendation in a 1942 publication that it
was “wise” to maintain hemoglobin levels “between 8 and 10
grams per cubic centimeter” for patients who were poor surgical risks by giving a preoperative transfusion (16). No data
were available to support this recommendation, but it stood
relatively unchallenged for about 50 years. An expanding body
of literature now suggests that arbitrary transfusion for a set
transfusion trigger (e.g., the “10/30 rule”) is ill-advised, and
that purported cardiac risks with anemia are overemphasized
(17). The following transfusion guidelines are presented based
on the best evidence currently available. Given the active ongoing investigations in this area, it is likely that frequent updates
will be forthcoming.
Whole Blood
There have been few widely accepted indications for whole
blood in modern transfusion practice. Storage of whole blood
precludes the extraction of components and, from a systems
perspective, is highly inefficient. As such, whole blood is not
available from most blood banks in the United States. In theory, the goals of oxygen delivery and volume expansion can be
achieved with packed RBCs and crystalloid solutions. Recent
experience with the use of whole blood by the U.S. military (18)
has rejuvenated the cause of whole blood. This accumulating
experience, especially with fresh whole blood having potentially beneficial effects on coagulopathy and hypothermia, may
result in some modification of civilian practices in the future.
Red Blood Cells
Packed RBCs are the most commonly utilized blood product,
providing oxygen-carrying capacity in cases of acute or chronic
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
blood loss. The longest storage life currently allowed by the
U.S. Food and Drug Administration (FDA) is 42 days. Longer
storage times result in fewer than 75% of the RBCs remaining viable in circulation 24 hours after transfusion. Platelets
degenerate at refrigerator temperatures, so refrigerated packed
RBCs contain essentially no functioning platelets. The levels of
factors V and VIII decrease significantly at 1◦ C to 6◦ C, while
levels of other factors remain essentially unchanged. There are
insignificant amounts of plasma in a unit of AS red cells.
Packed RBCs provide oxygen-carrying capacity and maintain oxygen delivery provided that intravascular volume and
cardiac function are adequate. The decision to transfuse, and
the amount of packed RBCs transfused, depend on the clinical situation. As noted previously, the use of a hematocrit of
30% (or a hemoglobin of 10 g/dL) as a transfusion trigger
is no longer acceptable. One or more units of blood may be
transfused with no predetermined number of units applicable.
Each unit of packed RBCs typically raises the hematocrit 2%
to 3% in a 70-kg adult, although this varies depending on the
donor, the recipient’s fluid status, the method of storage, and
its duration.
With blood loss, oxygen delivery is maintained through a
series of complex interactions and compensatory mechanisms.
This includes increased cardiac output, increased extraction ratio, rightward shift of the oxyhemoglobin curve, and expansion
of volume. Many anemic patients tolerate hemoglobin levels of
7 to 8 g/dL or less, as has been demonstrated in chronic renal
failure and Jehovah’s Witnesses (17). In general, cardiac output
does not increase significantly until hemoglobin falls below approximately 7 g/dL. Young healthy patients tolerate acute anemia to hemoglobin levels of 7 g/dL or less through increases
in cardiac output, provided they have a normal intravascular
volume and high arterial oxygen saturation.
In a multicenter, randomized controlled study of transfusion
in 838 patients in the critical care setting, a liberal transfusion strategy (transfusion for hemoglobin <10 g/dL) was compared with a restrictive strategy (transfusion for hemoglobin
<7 g/dL). The restrictive strategy was found to be at least as
effective as the liberal strategy, with the possible exception of
patients with acute myocardial infarction and unstable angina
(19). Suggested guidelines for RBC transfusion are listed in Table 171.1.
Leukocyte-reduced Red Blood Cells
The transfusion of RBCs has been associated with immunosuppression. This effect is thought to be related to exposure
to leukocytes. Therefore, the use of leukocyte-reduced components has been proposed as a means of minimizing immunosuppression; the majority of red cells and platelet transfusions
in the United States are currently leukocyte reduced. The efficacy of these preparations remains controversial, however,
TA B L E 1 7 1 . 1
Ongoing bleeding with hemodynamic instability
unresponsive (or incompletely responsive) to infusion of
2,000 to 3,000 mL crystalloid
■ Hemoglobin <7 g/dL
TA B L E 1 7 1 . 2
To decrease the incidence of subsequent refractoriness to
platelet transfusion caused by human leukocyte antigen
(HLA) alloimmunization in patients requiring long-term
platelet support
■ To provide blood components with reduced risk for
cytomegalovirus transmission
■ To prevent future febrile nonhemolytic transfusion
reactions (FNHTRs) in patients who have had a
documented FNHTR
■ To decrease the incidence of HLA alloimmunization in
nonhepatic solid-organ transplant candidates.
Data from Ratko TA, Cummings JP, Oberman HA, et al. Evidencebased recommendations for the use of WBC-reduced cellular blood
components. Transfusion. 2001;41:1310–1319.
and compelling data are lacking. Recent recommendations for
transfusion of leukocyte-reduced blood components are listed
in Table 171.2 (20).
Platelet transfusions are indicated for patients who are at a significant risk of bleeding because of quantitative or qualitative
platelet deficits. A unit of platelets can be prepared from individual (or “random”) donors or by apheresis, whereby a donor
provides the equivalent of 6 to 10 single “random” donor units.
In selected cases, human leukocyte antigen (HLA)-matched
platelets can be obtained by apheresis from HLA-matched
donors. The efficacy of platelet transfusion may be assessed
both by clinical parameters (improved hemostasis) and by following the platelet counts at 1 hour and 24 hours as an estimate
of platelet survival. The platelet count at 1 hour post transfusion of a unit of platelets should increase by 5,000 to 10,000
platelets/μL. Less pronounced responses should be expected
with repeated transfusion and the development of alloimmunization, or in the presence of fever, sepsis, or splenomegaly. If
alloimmunization is thought to be the cause of a poor response,
platelets from an HLA-matched donor may be needed.
The prophylactic transfusion of platelets in the absence of
microvascular bleeding, a low platelet count in a patient undergoing a surgical procedure, or a platelet count that has fallen
below 10,000 platelets/μL, in most medical patients, should
be considered inappropriate. Disease state–specific triggers for
platelet transfusion have been proposed and are listed in Table
171.3 (21). It is crucial to recognize that hypothermia depresses
platelet function, and platelet transfusion is generally ineffective with depressed temperatures. Restoration of a normal temperature returns platelet function to normal and ameliorates
microvascular bleeding.
Plasma is used as a source of clotting factors in patients with
coagulopathy and documented factor deficiency. This may occur with liver dysfunction, congenital absence of factors, and
transfusion of factor-deficient blood products, or after the use
of warfarin. A unit of plasma contains near-normal levels of all
factors, including about 400 mg of fibrinogen, and generally
TA B L E 1 7 1 . 5
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
TA B L E 1 7 1 . 3
Printer: Yet to come
Disseminated intravascular coagulation: 20,000–50,000
Major surgery in leukemia: 50,000 platelets/μL
Thrombocytopenia with massive transfusion: 50,000
Invasive procedures in cirrhosis: 50,000 platelets/μL
Cardiopulmonary bypass: 50,000–60,000 platelets/μL
Liver biopsy: 50,000–100,000 platelets/μL
Neurosurgical procedures: 100,000 platelets/μL
Data from Rebulla P. Platelet transfusion trigger in difficult patients.
Transfus Clin Biol. 2002;9:249–254.
Hemophilia A
von Willebrand disease
Uremic bleeding
As substitute for plasma if lower volume is desired
tions allow the use of a smaller volume of cryoprecipitate than
would be required if plasma were administered.
Risks of Blood Transfusion
increases factor levels by about 3%. Adequate clotting can usually be achieved with factor levels greater than 30%, although
higher levels are advisable in patients undergoing operative or
invasive procedures. The prothrombin time (PT) and the activated partial thromboplastin time (aPTT) can be used to assess
patients for plasma transfusion and to follow the efficacy of
administered plasma. Recent experience suggests that the use
of thromboelastography may provide advantages over the PT
as a guide for the treatment of coagulopathy (22,23). Plasma
can be frozen and stored for up to 1 year.
Plasma should not be given routinely or prophylactically
by “cookbook” formula after RBC transfusion—for example,
2 units of plasma for every 5 units of packed RBCs—or “prophylactically” after cardiac bypass or other procedures. Plasma
should not be used as a volume expander since crystalloids are
cheaper, safer, and at least as effective. Broadly accepted guidelines for transfusion of plasma are listed in Table 171.4 (24).
Indications for the use of cryoprecipitate include factor deficiency (hemophilia A), von Willebrand disease, and hypofibrinogenemia (Table 171.5). Some patients with uremic bleeding
may also benefit from cryoprecipitate transfusion. Cryoprecipitate is usually administered as a transfusion of 10 single units.
Each 5- to 15-mL unit contains over 80 units of factor VIII and
about 200 mg of fibrinogen. These relatively high concentra-
Even though a blood transfusion is a potentially life-saving
intervention, significant risks are still involved in the administration of these products. Risks range from minor febrile transfusion reactions to the transmission of viral infection to a potentially fatal transfusion of incompatible blood (Table 171.6).
Blood banks in the United States generally conduct over ten
individual tests or checks on donated units of blood in addition to the screening interview. Most (nine) are for infectious
diseases. Screening of donors and the introduction of increasingly effective tests for hepatitis and human immunodeficiency
TA B L E 1 7 1 . 6
TA B L E 1 7 1 . 4
International normalized ratio (INR) >1.5 with an
anticipated invasive procedure or surgery
Massive hemorrhage (over one blood volume) with an INR
Treatment of thrombotic thrombocytopenia purpura
Inherited coagulopathies where a specific factor
concentrate is not available
Emergent reversal of anticoagulant therapy
Data from Toy P, Popovsky MA, Abraham E, et al., and the National
Heart, Lung and Blood Institute Working Group on TRALI.
Transfusion-related acute lung injury: definition and review. Crit Care
Med. 2005;33:721–726.
Transfusion-related acute lung injury
Bacterial contamination of blood products
Administrative error leading to transfusion of ABOincompatible blood
Viral infection transmission
Hepatitis B
Hepatitis C
Human immunodeficiency virus 1 and 2
Human T-cell leukemia virus 1 and 2
Epstein-Barr virus
Parvovirus B19
Human herpesvirus 8
Transfusion-transmitted virus
Mad cow disease (bovine spongiform encephalopathy)
West Nile virus
Bacterial/protozoal infection transmission
Babesia microti
Trypanosoma cruzi
Yersinia enterocolitica
Serratia marcescens
Staphylococcus aureus
Staphylococcus epidermidis
Klebsiella pneumoniae
Trypanosoma cruzi
Transfusion reactions
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
West Nile
<-- CJD
General anesthesia
Metabolic risk in neanates
Under transfusion
FIGURE 171.1. Estimates of the current risk per unit of blood transfusion. The vertical bars represent log risk estimates (1–10, 1–100,
etc.). The dashed edges to lighter shaded horizontal bars signify that
the upper and lower estimates of risk are uncertain. CJD, CreutzfeldtJacob disease; HIV, human immunodeficiency virus; HCV, hepatitis C
virus; HBV, hepatitis B virus; TRALI, transfusion-related acute lung
injury; TA-GVHD, transfusion-associated graft versus host disease.
(From Dzik WH. Emily Cooley Lecture 2002: transfusion safety in
the hospital. Transfusion. 2003;43:1190–1199.)
virus (HIV) have dramatically reduced the risks of transmission
of these infections. The public has historically been most concerned about the transmission of HIV; however, recent data
reveal that the leading causes of fatalities after blood transfusion continue to be administrative error, leading to transfusion of ABO-incompatible blood, bacterial contamination, and
transfusion-related acute lung injury (TRALI). Overall, infectious risks of blood transfusion are far outweighed by noninfectious risks (Fig. 171.1). According to the FDA, TRALI
was the leading cause of transfusion-related mortality in 2003
(24). An average of 11.7 deaths from bacterial sepsis per year
in the United States was reported to the FDA from 2001 to
2003. This decreased to 7.5 deaths per year in 2004 and 2005,
due at least in part to the mandating of bacterial screening of
platelets, which began in 2004 (24,25). Transfusion of blood to
the wrong person continues to be a serious threat to patients.
In a review of a 10-year experience in New York State, Linden
et al. estimated the risk of an ABO-incompatible transfusion at
1 in 38,000 units of red cells, with the risk of a fatal reaction at
1 in 1.8 million transfusions (26). A rate of ABO-incompatible
transfusion of 1 in 12,000 units of red cells transfused has been
reported from the hemovigilance program in Quebec, Canada
Given the risks and benefits of blood transfusion, obtaining
informed consent for transfusion of blood components is crucial in nonemergent situations. A summary of the rates of the
more common, or more concerning, risks is provided in Table
171.7. These data may be useful for discussions with patients
and their families, and may be incorporated into informational
brochures addressing the risks and benefits of blood products
(Fig. 171.2).
Transfusion Reactions
The classification of the American Association of Blood Banks
for transfusion reactions is shown in Table 171.8. Hemolytic
transfusion reactions can be categorized broadly into acute
(<24 hours) and delayed (>24 hours) reactions. Hemolytic
TA B L E 1 7 1 . 7
Risk from blood product transfusion
Febrile nonhemolytic transfusion
Severe acute hemolytic reaction
Occurs in 0.5%–38% of all
Fatal in 1 of 600,000 transfusions
Delayed hemolytic reaction
1 in 260,000 transfusions
Bacterial contamination
1 in 15,000 platelet transfusions
Hepatitis B virus
Of concern in low-birth-weight
infants and immunocompromised
patients (e.g., transplant)
<1 in 137,000
Usually mild fever only
More common with platelet transfusions
Stop transfusion immediately and initiate
supportive measures
Suspect when unexplained fever, fall in
hematocrit, or jaundice occur
Among leading causes of transfusionrelated fatalities
Between 50% and 85% of adults in the United
States are carriers
Hepatitis C virus
<1 in 1 million
Human immunodeficiency virus
Human T-cell leukemia virus 1 and 2
West Nile virus
<1 in 1.9 million
Very small
Very small when donors are
properly screened
Transfusion-related lung injury
1 in 5,000 units (estimated)
25% of carriers have active hepatitis and may
progress to cirrhosis
Most infected persons asymptomatic, but 80%
become chronic
Potentially fatal
Rarely found in U.S. blood donors
80% of those infected remain asymptomatic,
20% develop mild symptoms, and 1 in
150–200 infected people develop severe disease
that may be fatal
5%–10% fatal
Source: American Association of Blood Banks. Blood/Facts About Blood and Blood Banking/fabloodtrans.htm. Accessed
Printer: Yet to come
Please realize you will be given blood or blood
products only if necessary. This brochure has
been offered to help you understand the benefits,
risks of, and alternatives to a blood transfusion
and is not inclusive of all information. Your
physician is the best source for additional
information related to blood transfusions.
Blood products could include:
• Red blood cells (known as “blood”)
• Platelets
• Fresh frozen plasma (known as “plasma”)
• Cryoprecipate (a specific part of plasma)
Blood products are prepared from
carefully screened, healthy, human
Any treatment in medicine involves weighing
the benefits and risks for each particular paitent.
In your case, your physician has recommended
that you receive a blood product. However, you
need to understand the potential complications
of transfusions, and also the consequences of
NOT receiving that blood product.
October 25, 2008
All blood products have a minimal risk of
transmitting an infectious disease.
All blood products are tested for transfusiontransmitted diseases according to federal
regulations. Yes, some risks remain present, as no
screen can be 100% effective. Viruses are not
commonly transmitted but may cause a serious
March 2004 estimates from the American
Association of Blood Banks show that the
risk of getting:
• HIV is <1 in 1.9 million transfused units
• Hepatitis C is <1 in million transfused
• Hepatitis B is <1 in 137,000 transfused
Source: About Blood
Other risks include, but are not limited to:
Bacterial contamination of a unit of blood
may occasionally occur and result in lifethreatening infections.
Transfusion errors, although rare, are also
a potential risk of blood transfusion.
Transfusion reactions, which are
unpredictable, may occur. Their symptoms
• Some reactions may present as fever and chills.
• Some reactions are mild immunologic reactions
that manifest 10–14 days after red cell
transfusion and may shorten the life of the
transfused red cells.
• Allergic reactions may be caused by plasma
proteins and typically cause itching and hives.
Some reactions may be serious.
• Serious reactions are extremely rare but may
include life-threatening disorders such as
shortness of breath or hemolysis (destruction of
the transfused red cells). This can in turn cause
jaundice or kidney problems.
Blood may be transfused to treat anemia or acute
blood loss.
• Anemia: A deficiency in the oxygen-carrying
material of the blood, also known as low blood count
or low blood level.
• Symptoms that might improve after blood
transfusion include weakness, shortness of breath,
chest pain, or light headedness.
• Conditions that may be prevented with appropriate
use of blood include strokes, heart attacks, kidney
failure, and other serious problems, including death.
• Platelets, plasma, or cryoprecipitate may be
prescribed for the prevention or treatment of
bleeding problems.
FIGURE 171.2. Example of patient and family information booklet for blood transfusion. (Courtesy of
Inova Fairfax Hospital, Falls Church, VA.)
TA B L E 1 7 1 . 8
Fever and/or chills, nonhemolytic
Transfusion-associated acute lung injury
Hypotension (associated with
angiotensin-converting enzyme
Circulatory overload
Nonimmune hemolysis
Air embolus
Transfusion-associated sepsis
Clinical features
Chills, fever, hypotension, renal failure, back
pain, hemoglobinuria
Temperature elevation >1◦ C, chills and/or
rigors, headache, vomiting
Pruritus, urticaria, flushing
Hypotension, urticaria, bronchospasm,
respiratory distress, wheezing, local edema,
Hypoxemia, respiratory failure, hypotension,
Caused by red blood cell mismatch
Varies from isolated urticaria to fatal
Leading cause of transfusion-associated
Flushing, hypotension
Dyspnea, orthopnea, cough, tachycardia,
hypertension, headache
Sudden dyspnea, cyanosis, chest pain, cough,
hypotension, cardiac arrhythmia
Paresthesia, tetany, arrhythmia
Cardiac arrhythmia
Bacterial contamination of transfused blood
Caused by physical destruction of blood
(heating, freezing, etc.)
Consider in patients with fever >40◦ C
and/or cardiovascular collapse
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
reactions occur when destruction of transfused RBCs occurs
because of preformed antibodies, and is mediated by complement. The reaction varies from mild to severe, depending
on the degree of complement activation and cytokine release
(28). Severe acute hemolytic reactions are usually due to the
transfusion of ABO-incompatible blood. Fatalities occur in 1
in 600,000 units, secondary to severe hemolytic transfusion
reactions (29). Red cell destruction results in the release of
peptides, which leads to hypotension, poor renal blood flow,
an activated coagulation cascade, and, in more severe cases,
disseminated intravascular coagulation (DIC). Clinical features
include pain and redness along the vein used for infusion, chest
pain, a feeling of doom, hypotension, oozing from wounds and
intravenous sites, chills, fever, oliguria, and hemoglobinuria. In
patients who are heavily sedated or unconscious, hypotension,
hemoglobinuria, and diffuse oozing suggest the diagnosis of
a severe, acute, hemolytic transfusion reaction. The diagnosis
may be especially difficult in the patient under general anesthesia, and a high index of suspicion is needed in order to make a
prompt diagnosis in such patients.
When a hemolytic or anaphylactic transfusion reaction is
suspected, the infusion should be stopped immediately and the
unit checked against the recipient’s identification band to determine whether the wrong unit has been administered to the
patient. The unit, including all intravenous solutions and tubing, should be sent promptly to the blood bank for examination. Blood should be drawn from a remote site and tested
for free hemoglobin. The urine should also be tested for free
hemoglobin. A direct antiglobulin test is indicated. Aggressive
fluid resuscitation should be initiated, and urine output should
be maintained at high levels. The early development of hypotension and DIC is associated with increased mortality.
Delayed hemolytic reactions tend to present 5 to 10 days
after transfusion (28), with approximately 1 in 260,000 patients developing a significant hemolytic reaction (30). The degree of hemolysis may be significant in the patient whose total
RBC mass has been replaced by massive transfusion. A transfused patient who develops an unexplained fall in hematocrit,
fever, or jaundice should be evaluated for the possibility of a
hemolytic reaction. The workup is similar to that for acute
hemolytic reactions, and the need for clinical intervention is
less likely.
Allergic nonhemolytic reactions are generally believed to be
caused by recipient antibodies to infusing donor plasma proteins. The manifestations vary from a slight rash or urticaria
to hemodynamic instability, with bronchospasm and anaphylaxis. Allergic reactions may be prevented by premedication
with antihistamines (e.g., diphenhydramine). Recipient antibodies against antigens on donor leukocytes or platelets will
produce febrile nonhemolytic reactions. Fevers and chills characterize these reactions shortly after the transfusion has started;
thus, an acute hemolytic reaction and bacterial contamination
of the unit should be ruled out. Treatment consists of antipyretics and transfusion of leukocyte-depleted blood components
when pharmacotherapy fails.
Hypocalcemia rarely occurs in patients receiving 1 unit of
blood at a time. The “prophylactic” use of calcium following blood transfusion is not evidence based. Patients receiving
large volumes of citrated blood, especially if their liver function is compromised and/or they are hypothermic (e.g., liver
transplant patients during their anhepatic phase in the operating room) are at greatest risk of hypocalcemia. In such unusual
situations, treatment with calcium gluconate (not calcium chloride) may be needed, and ionized calcium determinations can
help guide therapy.
Circulatory overload is a documented risk of blood transfusion, and thus, appropriate precautions should be taken in
patients with borderline cardiac and/or renal function and the
elderly. Particular care should be exercised in patients who require large volumes of plasma to rapidly correct coagulopathy associated with the use of warfarin. This should not be
interpreted to mean that every patient who receives a blood
transfusion should also be given a diuretic.
Transfusion-related Acute Lung Injury
The acute onset of pulmonary edema associated with transfusion and leading to death was first described in 1951 by Barnard
(31). The term, TRALI, was introduced by Popovsky in 1983
(32). It is currently the leading cause of death after transfusion,
with an estimated rate of 1 in 5,000 units transfused, although
higher rates have been reported (33). Transfusion-related acute
lung injury is likely underappreciated and underdiagnosed
due to other more commonly recognized conditions (such as
acute lung injury [ALI] and the acute respiratory distress syndrome [ARDS]) being often associated with blood transfusion, making the diagnosis of TRALI more difficult. The mortality rate associated with TRALI is in the range of 5% to
10% (34). These data suggest that all patients receiving blood
products should be appropriately monitored, including pulse
Transfusion-related acute lung injury occurs with the transfusion of all blood components, but especially platelets and
plasma. The clinical syndrome is characterized by the acute
onset of dyspnea, hypotension, hypoxemia, fever, and noncardiogenic pulmonary edema. The symptoms appear within 6
hours of transfusion, most often within 30 minutes. Since it
is similar to many other conditions encountered in the critical care setting, the diagnosis of TRALI is made by exclusion. Like other etiologies of acute lung injury, TRALI causes
an increase in pulmonary microvascular permeability with increased protein levels in the edema fluid. Two theories of the increased pulmonary microvascular permeability have been proposed in patients who develop TRALI. The first hypothesis
suggests that leukocyte antibodies from the donor unit activate recipient leukocytes in the pulmonary circulation, leading
to increased microvascular permeability and noncardiogenic
pulmonary edema. Blood donations from multiparous women
have been implicated as a contributing factor for TRALI, possibly because of increased leukocyte antibody levels. The second
hypothesis assumes an initial predisposing event that primes the
patient’s neutrophils and sequesters them in the lung. Biologically active lipids and cytokines in the donor unit then further
prime and activate the recipient’s neutrophils, with resultant
microvascular permeability and noncardiogenic pulmonary
The treatment of TRALI is supportive and consists of appropriate hemodynamic and ventilatory support. Once TRALI
is suspected, the transfusion should be terminated immediately
and the blood bank notified. The donor unit can be tested for
anti-HLA and/or antigranulocyte antibodies.
Transmission of Infection
Numerous viral and bacterial diseases may be transmitted by
blood transfusion (Table 171.6). Since March 1999, pooled
Printer: Yet to come
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
nucleic acid amplification testing (NAT) has been used to test
for HIV and hepatitis C virus (HCV), which involves pooling
of 16 to 24 individual blood samples and polymerase chain
reaction or other amplification techniques to test for HIV and
HCV nucleic sequences. Bacterial and protozoal diseases include syphilis, malaria, and infection with Babesia microti, Trypanosoma cruzi, Yersinia enterocolitica, Serratia marcescens,
Staphylococcus aureus, Staphylococcus epidermidis, or Klebsiella pneumoniae; Trypanosoma cruzi causes Chagas disease,
but transmission of this infection is very rare in the United
Bacterial Contamination
Bacterial contamination of blood is the most frequent cause of
transfusion-transmitted infectious disease (35). After hemolytic
reactions and TRALI, bacterial contamination is the most frequently reported cause of transfusion-related fatalities to the
FDA (36). The agents most often implicated in packed RBC
bacteremia were Serratia and Yersinia. For platelets, S. aureus,
Escherichia coli, Enterobacter, and Serratia species were more
frequently identified. Fever, chills, hypotension, tachycardia,
and shock after transfusion should raise the suspicion of bacterial contamination, and blood cultures of the patient and
unit should be obtained. Platelets, which are stored at 20◦ C
to 24◦ C, are a good growth medium for bacteria. Platelets
are now screened for bacterial contamination in the United
Transmission of the infectious agents for hepatitis is among
the most serious risks of blood transfusion. Past estimates of
posttransfusion hepatitis were approximately 10%. Current
data suggest that the infectious risk of hepatitis is <0.01%
per unit transfused (30). All blood is screened for the hepatitis
B virus (HBV), with tests for HBS Ag and anti-HBC . In addition, blood is screened for HCV with anti-HCV testing. The
risk of transfusion-associated HBV infection is approximately
1 in 30,000 to 1 in 250,000 per unit. With the development of
pooled NAT tests for HCV, the window period has decreased,
and the risk of HCV transmission is now as low as 1 in 1
million (37). No new case of transfusion-associated HCV has
been detected by the Centers for Disease Control and Prevention Sentinel Counties Viral Hepatitis Surveillance System since
1994 in the United States.
Approximately half of the blood recipients who contract
HBV infection develop symptoms; a much smaller percentage
requires hospitalization. Approximately half of patients who
contract posttransfusion HCV infection develop a chronic form
of the disease. Many of those patients eventually develop significant liver dysfunction, including cirrhosis.
Human Immunodeficiency Virus
The risk of HIV transmission from blood transfusion has decreased dramatically since the early 1980s despite an increasing incidence of HIV infection in the general population. The
window period from initial infection to the development of antibody to the virus poses a problem with the ability to detect
all seropositive donors. With pooled NAT, the window period
for detection of HIV has been reduced by 30% to 50%, and
the risk of HIV transmission is estimated to be as low as 1 in
2 million units (37).
Human T-cell Leukemia Virus
In addition to the transmission of cytomegalovirus (CMV),
hepatitis infection, and HIV, blood transfusion carries the risk
of transmission of human T-cell leukemia virus (HTLV) 1 and 2
infection. Transmission of the virus, especially to immunocompromised patients, may cause illnesses such as T-cell leukemia,
spastic paraparesis, and myelopathy, and has prompted routine
screening of donors in the United States since 1989. The risk
of HTLV 1 and 2 transmission is estimated to be 1 in 641,000
CMV infection is endemic, so routine screening is not performed in the United States. About 20% of blood donors are
infected with CMV by 20 years of age, and approximately
70% are infected by 70 years of age. The infection is carried in
white blood cells (WBCs). Most patients who encounter problems with CMV are immunocompromised, especially transplant recipients on immunosuppressive drugs. Such patients require transfusion with CMV-reduced-risk—leukocyte-reduced
or seronegative—blood products to avoid the transmission of
this viral infection. Human herpesvirus 8 causes Kaposi sarcoma and lymphoma in patients with acquired immunodeficiency syndrome (AIDS) and other immunosuppressed states.
Graft Versus Host Reaction
Blood transfusion exposes the recipient to many cells and proteins from the donor. When immunologically competent lymphocytes are introduced into an immunocompromised patient,
a graft versus host reaction can occur (28). The functional
donor lymphocytes attack recipient tissues, notably the bone
marrow, causing aplasia. Patients present with fever, rash, nausea, vomiting, diarrhea, liver function test abnormalities, and
depressed cell counts. This complication is fatal in as many as
90% of the cases. The prevalence of this complication in the
United States is not known but is thought to be rare. Rare cases
have also been reported from familial directed donations and
with HLA-matched platelets. γ -Irradiation of blood products
eliminates this risk.
Allogeneic blood transfusion may alter the immune response
in individuals and susceptibility to infection, tumor recurrence,
and reactivation of latent viruses. It has been known since 1974
that the transfusion of packed RBCs depresses the immune response in patients undergoing renal transplantation; however,
it is unclear to what extent these immunosuppressive effects
exist in other recipients. Contradictory evidence exists concerning increased infections in patients given allogeneic blood
transfusions. Similar controversy also exists regarding the exact relationship of blood transfusions to increased recurrence
of tumor and poor prognosis. Early studies on colorectal cancer showed decreased survival and increased tumor recurrence
in patients who were heavily transfused. Since then, studies
on many tumors have been performed that have not yielded
a decisive answer. The possibility exists that blood transfusion may represent a covariable, because very ill patients and
those undergoing more difficult procedures for more extensive
disease are more likely to receive blood transfusion. In light
of the immunomodulating effects of allogeneic blood transfusion, leukocyte-depleted transfusions have been suggested as an
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
alternative. In view of the data on immunosuppression from
blood transfusion, it would seem reasonable to adopt a policy of blood conservation in the perioperative period in the
absence of clear indications and acute symptoms. Leukocyte
reduction of blood products is thought to decrease the risk of
immunomodulation (38).
Decision Making in Blood Transfusion
Blood Transfusion in Hemorrhagic Shock
During World War I, it was believed that toxins caused vascular collapse in injured patients (39). Experiments in the 1930s
by Dallas B. Phemister and Alfred Blalock showed that fluid
was lost from the circulation into damaged tissues: the concept of fluid loss into a “third space.” During World War II,
plasma was the resuscitation solution of choice, as blood was
rarely available. British forces in the North African campaign
did utilize blood for casualties and noted improved outcome.
Although solutions containing electrolytes were used for children with diarrhea, and advances in research had increased the
understanding of metabolic and endocrine changes seen with
injury, the use of plasma solutions prevailed until the Korean
conflict. Subsequent experimental work indicated that extracellular fluids shifted into the intracellular space after significant
hemorrhage with shock (40). Providing volume resuscitation
in excess of shed blood became standard practice to maintain
adequate circulation and to refill the “third space.”
During World War II, acute tubular necrosis (ATN) was a
common consequence of hypovolemic shock. As fluid resuscitation became more prevalent during the Korean and Vietnam
conflicts, the incidence of ATN decreased. Yet, while posthypovolemic shock ATN became less common with better fluid
resuscitation, the acute—initially termed the adult, to differentiate it from the neonatal syndrome—respiratory distress syndrome became increasingly common. The lung injury in ARDS
was shown to be a function of the shock state rather than the
resuscitation solution used.
The goal of resuscitation from shock is prompt restoration
of adequate tissue and end-organ perfusion and oxygen transport. The American College of Surgeons Committee on Trauma
developed a classification of hemorrhagic shock that permits
useful guidelines for resuscitation (Table 171.9). Crystalloid is
infused at a 3:1 ratio for every unit of RBCs administered, and
therapy is monitored primarily by hemodynamic response. Because crystalloid solutions are universally available, and some
delay is required to prepare blood products, crystalloid is the
proper initial resuscitation fluid. Resuscitation proceeds with
the use of blood products, depending on the patient’s response.
Although controversy existed in the past regarding the choice
of a colloid solution (e.g., albumin, plasma) or a crystalloid
solution (e.g., lactated Ringer [LR] solution or saline), recent
evidence has confirmed that colloid solutions offer no advantages over crystalloids for fluid resuscitation in critically ill patients (41). Crystalloid solutions should be considered the solutions of choice because they are less expensive, need not be
cross-matched, do not transmit disease, and probably result in
less fluid accumulation in the lung. No experimental data indicate that using colloid rather than crystalloid solutions can
prevent pulmonary edema. An updated review of randomized
controlled trials of albumin resuscitation yielded no suggestion of a reduction in mortality when the colloid was used in
hypovolemia or in critically ill patients with burns and hypoalbuminemia (42).
Several crystalloid solutions are available for resuscitation,
but isotonic solutions should be used to avoid free water overload. While lactated Ringer solution is recommended as initial
therapy, metabolic alkalosis is common after successful resuscitation with this solution and blood products because the lactate in LR solution and the citrate in banked blood are both
converted to bicarbonate in the liver. LR solution contains calcium and, if it is mixed with a blood product, the blood may,
in theory, clot in the bag. Normal (0.9%) saline solution is
an acceptable alternative to LR solution, but large volumes
can produce a hyperchloremic metabolic acidosis, which may
complicate the use of base deficit in resuscitation. Since normal
saline is compatible with all blood products, its use is sometimes preferred if transfusion is a possibility.
The decision to transfuse blood is highly dependent on the
acuity of blood loss. Patients with acute, massive hemorrhage,
such as those with trauma or gastrointestinal bleeding, show
signs of hemodynamic instability early in their presentation.
The clinical picture depends on the amount of blood loss (Table 171.9). For example, acute loss of 40% of the total blood
volume (about 2,000 mL in a 70-kg patient) is associated with
severe tachycardia, hypotension, depressed mental status, and
TA B L E 1 7 1 . 9
Class of hemorrhage
Blood volume loss
15% (750 mL)
(750–1,500 mL)
(1,500–2,000 mL)
No resuscitation generally needed
Crystalloid resuscitation needed. Blood
transfusion given if no response to fluids (or
if response is transient)
Blood transfusion generally needed, with
crystalloids in 3:1 ratio
>40% (>2,000 mL)
Vital signs essentially normal
Tachycardia, decreased pulse
pressure, anxiety, pallor,
diaphoresis, acidosis
Hypotension, tachycardia,
decreased mental status,
Severe tachycardia and
hypotension, lethargy
Adapted from the American College of Surgeons, Advanced Trauma Life Support.
Massive resuscitation with fluids and blood
products needed
Printer: Yet to come
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
oliguria. On the other hand, blood loss of up to 15% of the
blood volume (750 mL) may not have any obvious physiologic
It is important to remember that the diagnosis of hemorrhagic shock and the decision to administer blood transfusion
should not be based solely on hypotension, tachycardia, or
anemia. Hypotension does not generally occur until more than
30% of the blood volume has been lost. This is particularly the
case in children who, due to very effective compensatory mechanisms, maintain their blood pressure despite severe blood loss.
Conversely, elderly patients on β-blocking agents may not manifest significant tachycardia. Hemoglobin levels obtained early
in the course of hemorrhagic shock do not reflect the severity
of blood loss, as there has not been enough time for fluid shifts
to occur. Therefore, blood transfusion should be based on a
comprehensive assessment of the patient, including vital signs
and estimation of the amount of blood loss, as well as clinical
and laboratory evaluation of end-organ perfusion.
Acute, massive hemorrhage is managed initially with aggressive volume replacement using crystalloid solutions. After administering 2,000 to 3,000 mL of crystalloid solution,
blood transfusion should be initiated in patients who continue
to manifest unstable vital signs. This should occur concomitantly with expeditious surgical control of the bleeding sites.
Cross-matched blood should be given as soon as it is available.
If needed, type O negative blood can be given to women of
childbearing age, and type O positive blood can be given to
men of all ages and women older than 50 years of age until
cross-matched blood is available. Correction of coagulopathy
and hypothermia is paramount. A “damage control” surgical
approach, aimed at rapid control of bleeding while delaying less
urgent procedures, should be utilized. This helps reduce transfusion requirements and allows the patient to recover more
quickly from shock.
stable angina and acute myocardial infarction, were excluded
from the study. In the latter group of patients, maintaining the
hemoglobin at or above 10 g/dL remains the standard of care,
although there are conflicting data on that subject.
In a study of Medicare discharge records, elderly patients
with acute myocardial infarction had a lower mortality if their
hematocrit was 30% or higher (54). Another study suggested
that a higher hematocrit upon admission to the ICU after coronary artery bypass grafting was associated with a higher rate of
myocardial infarction (55). Despite the large body of evidence
against empiric blood transfusion in normovolemic patients,
physicians continue to transfuse patients with hematocrit levels between 21% and 30% (56). Finally, there have also been
reports advocating a hematocrit level of 30% in septic patients
(57), although these reports do not establish blood transfusion
to a hematocrit of 30% as an independent factor contributing
to improved outcome. The general trend, overall, appears to be
that of an increasingly restrictive strategy of blood transfusion
Immediate Concerns
Given the known risks and the costs associated with blood
transfusions, a comprehensive strategy of blood conservation
should be followed. The need to correct anemia should be assessed, sources of ongoing blood loss should be controlled, and
measures to enhance erythropoiesis should be entertained.
Minimizing Unnecessary Blood Loss
Blood Transfusion in the Normovolemic Patient
Anemic patients with a normal blood volume, such as patients
who have recovered from hemorrhagic shock and those with
subacute or chronic anemia, are generally hemodynamically
intact. Concerns regarding the diminished oxygen-carrying capacity of the blood may persist in some of these patients, especially those in the critical care setting. For many years, the
standard of care dictated that a hematocrit level of at least 30%
should be maintained; the rationale included faster recovery
and prevention of myocardial ischemia, especially in patients
with coronary artery disease. Recent data indicate that lower
hematocrit levels are well tolerated, even in patients at risk for
myocardial ischemia (10,43–47). Combined with the current
understanding of blood transfusion risks, this has resulted in
lowering the trigger level for transfusion.
There are now many reports demonstrating that blood
transfusion is an independent risk factor for worse outcome, including increased mortality, especially in trauma patients (48–
53). In a landmark study, H´ebert et al. demonstrated that maintaining the hemoglobin at or above 10 g/dL (liberal strategy)
in euvolemic critically ill patients—as compared to maintaining the hemoglobin at 7 g/dL, the conservative strategy—was
not associated with any improvement in overall mortality (19).
In fact, mortality was significantly lower with the conservative
strategy (hemoglobin at 7 g/dL) among patients who were less
acutely ill and in those who were younger than 55 years of
age. Patients with active myocardial ischemia, defined as un-
A significant amount of blood can be lost with repeated phlebotomy in the ICU. This is particularly significant in children.
Routine serial “blood draws” should be avoided. A policy
of obtaining laboratory results only when clinically indicated
should be followed. Microsampling techniques, including bedside point-of-care testing, limit the amount of blood lost with
each blood draw. Since the estimated daily blood loss from
phlebotomy is at least 40 mL/day (10,11,13), critical care practitioners should carefully consider the need for frequent phlebotomy in the ICU.
Optimization of Red Cell Production
Iron is essential for properly functioning hemoglobin, as it is
the site of attachment of the oxygen molecule. Other oxygencarrying proteins, such as myoglobin and cytochrome a-a3,
also depend on iron. Many enzymes in the Kreb cycle contain
iron in their functional groups. In the critically ill patient, iron
deficiency anemia may be multifactorial, for example, poor gastrointestinal absorption, nutrient antagonism, and concomitant copper and vitamin A deficiencies (59).
Patients with the systemic inflammatory response syndrome
(SIRS) have circulating cytokines that impair the release of iron
stored in the reticuloendothelial system. This creates a situation
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
where total body iron levels are normal but iron is not available for incorporation into red cell precursors (functional iron
deficiency anemia).
Despite the central role that iron plays in oxygen delivery,
it is still not known whether iron supplementation in critically
ill anemic patients is beneficial (60). Perceived iron deficiency
could be functional, rather than an absolute reduction in total
body iron (61,62). In addition, iron supplementation has been
implicated with an increased risk and severity of infection since
free iron acts as a chelator of free radicals (63). There is currently no clear indication to administer supplemental iron to
critically ill patients who are anemic.
Adverse effects potentially attributable to erythropoietin
therapy include hypertension, thrombotic complications, cardiovascular events, tumor progression in cancer patients, and
increased risk of death. In November 2006, the FDA issued
an alert to provide new safety information for erythropoiesisstimulating agents (ESAs) (81). The alert was based on analyses
of studies on cancer and orthopedic surgery patients who were
found to have a higher chance of serious and life-threatening
effects and/or death with the use of ESAs. The FDA recommends using the lowest dose possible to achieve a hemoglobin
level that avoids the need for transfusion, and withholding the
dose of the ESA if the hemoglobin level exceeds 12 g/dL or rises
by 1 g/dL in any 2-week period.
Erythropoietin is a circulating glycoprotein secreted primarily
by the kidneys in response to hypoxia. Its principal action is to
stimulate the production and release of RBCs from the bone
marrow (64). This hormone is now commercially available using recombinant DNA technology, and has been approved for
use in anemic patients with end-stage renal disease. Its indications were extended to include anemic patients with chronic
renal insufficiency, cancer, and AIDS. The indications for erythropoietin therapy are still being expanded. Patients undergoing elective surgical procedures that are typically associated
with severe blood loss may benefit from preoperative erythropoietin therapy combined with autologous blood transfusion
The potential therapeutic value of erythropoietin in anemia
of critical illness is an area of intense research. Erythropoiesis
in critically ill patients can be suppressed for a variety of reasons, including renal and hepatic failure. Circulating cytokines
in SIRS suppress erythropoiesis both by blunting the response
to and inhibiting the production of erythropoietin (66–72).
Gabriel et al. noted that erythropoietin formation in patients
with multiple organ dysfunction was inadequate to stimulate
reticulocytosis in what was described as a relative erythropoietin deficit (73). In their study, high doses of recombinant human erythropoietin therapy did stimulate the erythropoietic
system, as evidenced by a higher rate of reticulocytosis. There
was, however, no increase in hematocrit or reduction in packed
RBC transfusion during the 3 weeks of the study.
Studies have focused on the potential of human recombinant
erythropoietin therapy to reduce transfusion requirements and
improve outcome in critically ill patients (74). Corwin et al.,
in two randomized controlled trials (75,76), demonstrated a
reduction of up to 19% in packed RBC units transfused and
a greater increase in hematocrit in the group treated with erythropoietin; there were no differences in morbidity or mortality
between the two groups. Georgopoulos et al. showed similar
results, with the additional finding that the effects of erythropoietin therapy are dose dependent (77).
More recent studies have shown less favorable results. Another study by Corwin et al. noted that the use of erythropoietin
alfa did not reduce the incidence of red cell transfusion among
critically ill patients, and treatment with this agent was associated with an increase in the incidence of thrombotic events (78).
A second study concluded that the use of a target hemoglobin
level of 13.5 g/dL in chronic kidney disease was associated with
increased risk and no improvement in quality of life (79). Several reports have also demonstrated adverse outcomes in cancer
patients (80).
Blood lost during surgical procedures can be retrieved, spun,
washed, and filtered. The recovered RBCs are then reinfused
back into the patient. Similarly, blood from drains such as thoracostomy tubes can be retrieved, collected in containers with
citrate solutions to prevent clotting, and reinfused. Relative
contraindications include contamination of blood with bacteria, malignant cells, or amniotic or ascitic fluids. Other strategies of blood conservation include preoperative autologous donation and acute normovolemic hemodilution (30).
Hemoglobin-based Oxygen Carriers
The search for a solution that can transport oxygen from the
lungs to the tissues started in the early part of the 20th century
and continues to the present day (82–84). These solutions are
loosely termed “blood substitutes,” although they should be
more appropriately described as “oxygen carriers,” since this
is the only blood function for which they substitute. A variety
of substances have been studied, including perfluorocarbons
and porphyrins. Research on the latter two categories of oxygen carriers has been largely abandoned due to problems with
manufacturing, ease of use, and adverse effects (85–87).
Current investigation is now focused on the hemoglobinbased oxygen carriers (HBOCs). Hemoglobin can be obtained
from three sources: human blood from discarded units of
packed RBCs, animal blood, and recombinant DNA technology.
Structure and Function of Normal
Human Hemoglobin
Hemoglobin (Hb) is a large molecule made up of four polypeptide chains (two α- and two β-chains), with a molecular weight
of 64,450. Each chain is conjugated with a heme moiety,
an iron-containing porphyrin derivative to which oxygen attaches, forming oxyhemoglobin. When fully saturated, each
Hb molecule has four oxygen molecules attached. Iron has to
be in the ferrous state (Fe2+ ) in order for oxygen to attach.
When blood is exposed to various drugs and other oxidizing
agents, ferrous iron is converted to ferric iron (Fe3+ ), forming methemoglobin (met-Hb), which cannot bind oxygen. An
enzyme within red cells, met-Hb reductase, converts met-Hb
back to Hb.
The affinity of Hb for oxygen increases exponentially as
more oxygen molecules attach, and hence the sigmoid nature
Printer: Yet to come
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
of the oxygen–Hb dissociation curve. Factors that decrease the
affinity of Hb to oxygen (i.e., making off-loading of oxygen
easier) include acidosis and 2,3-DPG.
Characteristics of Cell-free Hemoglobin
Dissociation. When free in the plasma, the Hb tetramer dissociates into two αβ-dimers, which are filtered through renal
glomeruli and can then precipitate in the renal tubules, causing
obstruction. This adverse effect is further compounded by the
decreased renal blood flow that results from the vasoconstrictive effect of Hb (88,89). Technologies were developed to produce large stable Hb polymers by cross-linking Hb molecules;
the most commonly used cross-linking reagent is glutaraldehyde. This process results in the formation of polymers of
varying sizes that do not filter through the glomeruli. Another strategy used to stabilize Hb was intramolecular crosslinking, whereby the cross-link was between α-chains of the
same molecule so that neither polymerization nor subunit dissociation occurred; this product was abandoned due to intense
vasoconstrictive features.
Viscosity. The lower viscosity of Hb solutions, compared to
blood, was initially thought to be advantageous, as it provided
less systemic vascular resistance. However, deeper insight into
the physiology of the vascular endothelium revealed that the
reduced shear stresses on the blood vessel wall were associated
with decreased secretion of relaxing factors such as prostacyclin and endothelin, with a net vasoconstrictive effect. The resulting decrease in blood flow antagonizes the oxygen delivery
function of Hb (90,91).
Vasoactivity. Most HBOCs have a systemic pressor effect
(92,93), and some have the same effect on the pulmonary circulation as well (94). In addition to the above mechanisms of
vasoconstriction, two other mechanisms are described: binding of nitric oxide and stimulation of catecholamine release;
these effects have been associated with decreased cardiac output (95).
Affinity for Oxygen. Once released from the red cell, Hb loses
its 2,3-DPG, and its affinity for oxygen increases. This causes
a leftward shift of the oxygen–Hb dissociation curve, thus impairing the off-loading of oxygen. Strategies to decrease the
affinity of Hb for oxygen include pyridoxalation and the use
of bovine Hb. It is not clear whether decreasing the affinity of
Hb for oxygen is beneficial. For example, higher levels of oxygen at the tissue level may trigger an autoregulatory response
by the blood vessel wall, whereby there is decreased secretion
of relaxing factors, resulting in vasoconstriction and decreased
flow (96).
Oxidation. Deprived of the met-Hb reductase in red cells, free
Hb is at higher risk of being oxidized into met-Hb. However, other antioxidants such as glutathione are available in
the plasma to serve this function. Levels of met-Hb in patients
receiving HBOCs do not appear to be physiologically significant (97).
Effects on the Inflammatory Response. HBOCs, unlike stored
blood, lack the ability to stimulate neutrophils and incite an
inflammatory response with its attendant systemic manifestations of multiple organ dysfunction (98).
Clinical Trials of Hemoglobin-based
Oxygen Carriers
The most widely studied HBOC in clinical practice is a human
polymerized hemoglobin product (PolyHeme, Northfield Laboratories, Evanston, IL). The first randomized trial in acute
trauma and emergency surgery was published in 1998 (99),
showing that PolyHeme maintained total hemoglobin in lieu
of red cells despite the marked fall in RBC hemoglobin, and reduced the use of blood transfusion. The study concluded that
PolyHeme appears to be a clinically useful blood substitute. A
phase III trial involving 720 patients from 32 level I trauma
centers was recently completed. The trial randomized trauma
patients with evidence of hemorrhagic shock at the scene to either normal saline or PolyHeme. Treatment was started in the
field and continued for up to 12 hours after injury. The primary
end point was survival at 30 days. Preliminary results showed
no statistically significant difference in survival between patients receiving PolyHeme without blood for up to 12 hours
following injury and those receiving the standard of care, including early blood replacement. PolyHeme may, therefore, be
useful when blood is needed but not available (100).
Critically ill patients with transfusion preferences present a
challenging management problem. For example, Jehovah’s
Witnesses’ refusal of blood and blood products is part of
their religious beliefs (Genesis 9:3-4, Leviticus 17:10-11) (101).
Honoring these beliefs requires modification of medical management strategies, and presents a unique opportunity to question transfusion guidelines and thresholds. The care of these
patients requires early identification of transfusion preferences.
All patients admitted to the critical care setting should have
treatment preferences (including blood transfusion) discussed
with them or their legal representative as soon as possible. Although transfusion may need to be administered in some emergent situations without the opportunity to obtain informed
consent, in most circumstances the critical care practitioner
should be able to discuss the risks, benefits, and potential
complications of transfusion of various blood products with
the patient or representative. Moreover, individual patients
may have preferences—religious or otherwise—regarding some
blood products but not others, so it is important to establish
these preferences for each blood product available. Discussion
with patients and family members should include a detailed
explanation of each blood product, as the origin and technical
aspects of these products may affect their acceptance. In the
case of the Jehovah’s Witness, or other groups with religious
preferences, assistance from a church representative or other
religious leaders may be extremely helpful to the family and
the physician.
Although survival at lower levels of hemoglobin (>3 g/dL)
have been reported, mortality rates exceed 50% when levels
fall below 3 g/dL (102). A recent experience with an injured
patient who was a Jehovah’s Witness demonstrated that survival without neurologic impairment was possible even at extremely low hemoglobin and hematocrit levels (2.7 g/dL and
7.8%, respectively) (103). The implementation of blood conservation strategies, hormonal stimulation, and the use of red
Printer: Yet to come
October 25, 2008
Chapter 171: Transfusion Therapy: When to Use It and How to Minimize It
TA B L E 1 7 1 . 1 0
Blood conservation
No routine blood draws
Consequential blood tests only
Use of capillary tubes for arterial blood gas analysis
Pediatric-size tubes for other tests
Decisive surgical interventions in cases of bleeding
Intraoperative and peri-procedure blood conservation
■ Hemostasis
■ Autologous transfusion (cell salvage device)
■ Normovolemic hemodilution
■ Maximization of oxygen delivery
Maintain high oxygen saturation
Minimize oxygen demand
■ Sedation
■ Mechanical ventilation
■ Neuromuscular blockade
■ Allow permissive hypercapnia/metabolic acidosis
■ Hormonal stimulation
High-dose recombinant erythropoietin
Iron supplementation
■ Red cell substitutes (as they become available)
cell substitutes as they become available are options in the management of these patients. The use of high-dose erythropoietin
(40,000 units subcutaneously every other day) and supplemental iron provide accelerated erythropoiesis under extreme circumstances. Table 171.10 lists potential strategies that may be
useful in the management of the Jehovah’s Witness and others who request that blood transfusion not be administered. A
number of these strategies should be considered for all patients
in the critical care setting to minimize the need for transfusion.
1. Webster C. The origins of blood transfusions: a reassessment. Med Hist.
2. Maluf NSR. A history of blood transfusion. J Hist Med. 1954;9:59–107.
3. Schmidt PJ. Transfusion in America in the eighteenth and nineteenth centuries. N Engl J Med. 1968;279:1319.
4. Blundell J. Successful case of transfusion. Lancet. 1828;1:431.
5. http ://˙Blood/Highlights of Transfusion
Medicine History/highlights.htm. Accessed January 5, 2008.
6. American Association of Blood Banks for the United States Department of Health and Human Services. 2005 Nationwide Blood Collection
and Utilization Survey Report.
and Services/Data Center/NBCUS/nbcus.htm. Accessed January 8, 2008.
7. Report on blood collection and transfusion in the United States in 2001.
National Blood Data Resource Center; 2003.
8. Goodnough LT, Brecher ME, Kanter MH, et al. Blood transfusion. N Engl
J Med. 1999;340:438–447, 525–533.
9. Cobain TJ, Vamvakas EC, Wells A, et al. A survey of the demographics of
blood use. Transfus Med. 2007;17:1–15.
10. Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion in the ICU. Is
there a reason? Chest. 1995;108:767–771.
11. Vincent JL, Baron J-F, Reinhart K, et al. Anemia and blood transfusion in
critically ill patients. JAMA. 2002;288:1499–1507.
12. Corwin HL, Gettinger A, Pearl RG, et al. The CRIT study: anemia and
blood transfusion in the critically ill—current clinical practice in the United
States. Crit Care Med. 2004;32:39–52.
13. Nguyen Ba V, Peres Bota D, Melot C, et al. Time course of hemoglobin
concentrations in non-bleeding ICU patients. Crit Care Med. 2003;31:406–
14. Rutledge R, Sheldon GF, Collins ML. Massive transfusion. Crit Care Clin.
15. Vincent JL, Piagnerelli M. Transfusion in the intensive care unit. Crit Care
Med. 2006;34(Suppl):S96–S101.
16. Adams RC, Lundy JS. Anesthesia in cases of poor surgical risk. Surg Gynecol Obstet. 1942;74:1011–1019.
17. Fakhry SM, Fata P. How low is too low? Cardiac Risks with Anemia. Crit
Care. 2004;8(Suppl 2):S11–S14.
18. Holcomb JB. Damage control resuscitation. J Trauma. 2007;62:S36–S37.
19. Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized,
controlled clinical trial of transfusion requirements in critical care. N Engl
J Med. 1999;340:409–417.
20. Ratko TA, Cummings JP, Oberman HA, et al. Evidence-based recommendations for the use of WBC-reduced cellular blood components. Transfusion. 2001;41:1310–1319.
21. Rebulla P. Platelet transfusion trigger in difficult patients. Transfus Clin
Biol. 2002;9:249–254.
22. Kheirabadi BS, Crissey JM, Deguzman R, et al. In vivo bleeding time
and in vitro thrombelastography measurements are better indicators of
dilutional hypothermic coagulopathy than prothrombin time. J Trauma.
23. Clark P, Mintz PD. Transfusion triggers for blood components. Curr Opin
Hematol. 2001;8:387–391.
24. Toy P, Popovsky MA, Abraham E, et al., and the National Heart, Lung and
Blood Institute Working Group on TRALI. Transfusion-related acute lung
injury: definition and review. Crit Care Med. 2005;33:721–726.
25. Blajchman MA, Vamvakas EC. The continuing risk of transfusiontransmitted infections. N Engl J Med. 2006;355:1303–1305.
26. Linden JV, Wagner K, Voytovich AE, et al. Transfusion errors in New
York State: an analysis of 10 years’ experience. Transfusion. 2000;40:1207–
27. Robillard P, Itaj NK, Corriveau P. ABO incompatible transfusions, acute
and delayed hemolytic transfusion reactions in the Quebec hemovigilance
system—year 2000 (abstract). Transfusion. 2002;42(suppl):25S.
28. Snyder E. Transfusion reactions. In: Hoffman R, Benz EJ, Shattil SJ, et al.,
eds. Hematology: Basic Principles and Practice. 3rd ed. Philadelphia:
Churchill Livingstone; 2000.
29. Goodnough LT. Erythropoietin therapy versus red cell transfusion. Curr
Opin Hematol. 2001;8:405–410.
30. Goodnough LT, Brecher ME, Kanter MH, et al. Blood transfusion. N Engl
J Med. 1999;340:438–447, 525–533.
31. Barnard R. Indiscriminate transfusion: a critique of case reports illustrating
hypersensitivity reactions. N Y State J Med. 1951;51:2399–2402.
32. Popovsky MA, Abel MD, Moore SB. Transfusion-related acute lung injury associated with passive transfer of anti-leukocyte antibodies. Am Rev
Respir Dis. 1983;128:185–189.
33. Toy P, Popovsky MA, Abraham E, et al., and the National Heart, Lung and
Blood Institute Working Group on TRALI. Transfusion-related acute lung
injury: definition and review. Crit Care Med. 2005;33:721–726.
34. Shander A, Popovsky MA. Understanding the consequences of transfusionrelated acute lung injury. Chest. 2005;128:598S–604S.
35. Reading FC, Brecher ME. Transfusion-related bacterial sepsis. Curr Opin
Hematol. 2001;8:380–386.
36. Kuehnert MJ, Roth VR, Haley NR, et al. Transfusion-transmitted bacterial infection in the United States, 1998 through 2000. Transfusion. 2001;
37. Strong DM. Infectious risks of blood transfusions. Am Blood Centers Blood
Bull. 2001;4(2).
38. Blajchman MA. Immunomodulation and blood transfusion. Am J Ther.
39. Rutherford EJ, Skeete D, Schooler WG, et al. Hematologic principles in
surgery. In: Townsend CM, Beauchamp RD, Evers BM, et al., eds. Sabiston
Textbook of Surgery: The Biologic Basis of Modern Surgical Practice. 17th
ed. Philadelphia: Elsevier-Saunders; 2004:113–136.
40. Shires T, Coln D, Carrico J, et al. Fluid therapy in hemorrhagic shock. Arch
Surg. 1964;88:688–693.
41. The SAFE Investigators. A comparison of albumin and saline for fluid
resuscitation in the intensive care unit. N Engl J Med. 2004;350:2247–
42. Alderson P, Bunn F, Lefebvre C, et al. Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane Database
Syst Rev. 2004;4:CD001208.
43. Hebert PC, Wells G, Berlin JA, et al. A Canadian survey of transfusion
practices in critically ill patients. Crit Care Med. 1998;26:482–487.
44. Carson JL, Willett LR. Is a hemoglobin of 10 g/dl required for surgery?
Med Clin North Am. 1993;77:335–347.
45. Sehgal LR, Zebala LP, Takagi I, et al. Evaluation of oxygen extraction ratio
as a physiologic transfusion trigger in coronary artery bypass graft surgery
patients. Transfusion. 2001;41:591–595.
46. Stehling L, Simon TL. The red blood cell transfusion trigger. Physiology
and clinical studies. Arch Pathol Lab Med. 1994;118:429–434.
47. Greenburg AG. A physiologic basis for red blood cell transfusion decisions.
Am J Surg. 1995;170:44S–48S.
48. Malone DL, Dunne J, Tracy JK, et al. Blood transfusion, independent of
shock severity, is associated with worse outcome in trauma. J Trauma.
49. Robinson WP, Ahn J, Stiffler A, et al. Blood transfusion is an independent
Printer: Yet to come
October 25, 2008
Section XIX: Hematologic and Oncologic Disease and Dysfunction
predictor of increased mortality in nonoperatively managed blunt hepatic
and splenic injuries. J Trauma 2005;58(3):437–445.
Hill GE, Frawley WH, Griffith KE, et al. Allogeneic blood transfusion
increases the risk of postoperative bacterial infection: a meta-analysis. J
Trauma. 2003;54(5):908–914.
Offner PJ, Moore EE, Biffl WL, et al. Increased rate of infection associated
with transfusion of old blood after severe injury. Arch Surg. 2002;137:711–
Zallen G, Offner PJ, Moore EE, et al. Age of transfused blood is an independent risk factor for postinjury multiple organ failure. Am J Surg. 1999;
Claridge JA, Sawyer RG, Schulman AM, et al. Blood transfusions correlate
with infections in trauma patients in a dose-dependent manner. Am Surg.
Wu W, Rathore S, Wang Y, et al. Blood transfusion in elderly patients with
acute myocardial infarction. N Engl J Med. 2002;345:1230–1236.
Spiess BD, Ley C, Body SC, et al. Hematocrit value on intensive care unit
entry influences the frequency of Q-wave myocardial infarction after coronary artery bypass grafting. The Institutions of the Multicenter Study of Perioperative Ischemia (McSPI) Research Group. J Thorac Cardiovasc Surg.
Shapiro MJ, Gettinger A, Corwin HL, et al. Anemia and blood transfusion in trauma patients admitted to the intensive care unit. J Trauma.
Rivers E, Nguyen B, Havstad S, et al. Early goal directed therapy in the
treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–
Earley AS, Gracias VH, Haut E, et al. Anemia management program reduces transfusion volumes, incidence of ventilator-associated pneumonia,
and cost in trauma patients. J Trauma. 2006;61(1):1–7.
Lynch SR. Interaction of iron with other nutrients. Nutr Rev. 1997;55:102–
Prelack K, Sheridan RL. Micronutrient supplementation in the critically
ill patient: strategies for clinical practice. J Trauma. 2001;51(3):601–
Deitch EA, Sittig KM. A serial study of the erythropoietic response to thermal injury. Ann Surg. 1993;217:293–299.
Krantz SB. Pathogenesis and treatment of the anemia of chronic disease.
Am J Med Sci. 1994;307:353–359.
Walter T, Olivares M, Pizarro F, et al. Iron, anemia, and infection. Nutr
Rev. 1997;55:111–124.
Jelkmann W. Erythropoietin: structure, control of production, and function.
Physiol Rev. 1992;72:449–489.
Goodnough LT, Monk TG, Andriole GL. Current concepts: erythropoietin
therapy. N Engl J Med. 1997;336:933–938.
Means R Jr, Krantz SB. Progress in understanding the pathogenesis of the
anemia of chronic disease. Blood. 1992;80:1639–1647.
Fuchs D, Hausen A, Reibnegger G, et al. Immune activation and the
anaemia associated with chronic inflammatory disorders. Eur J Haematol.
Baer AN, Dessypris EN, Goldwasser E, et al. Blunted erythropoietin response to anaemia in rheumatoid arthritis. Br J Haematol. 1987;66:559–
Faquin WC, Schneider TJ, Goldberg MA. Effect of inflammatory cytokines
on hypoxia-induced erythropoietin production. Blood. 1992;79:1987–
Jelkmann W, Pagel H, Wolff M, et al. Monokines inhibiting erythropoietin production in human hepatoma cultures and in isolated perfused rat
kidneys. Life Sci. 1992;50:301–308.
von Ahsen N, Muller C, Serke S, et al. Important role of nondiagnostic
blood loss and blunted erythropoietic response in the anemia of medical
intensive care patients. Crit Care Med. 1999;27:2630–2639.
Rogiers P, Zhang H, Leeman M, et al. Erythropoietin response is blunted
in critically ill patients. Intensive Care Med. 1997;23:159–162.
Gabriel A, Kozek S, Chiari A, et al. High-dose recombinant human erythropoietin stimulates reticulocyte production in patients with multiple organ
dysfunction syndrome. J Trauma. 1998;44(2):361–367.
Corwin HL. The role of erythropoietin therapy in the critically ill. Transfus
Med Rev. 2006;20:27–33.
Corwin HL, Gettinger A, Rodriguez RM, et al. Efficacy of recombinant
human erythropoietin in the critically ill patient: a randomized, doubleblind, placebo-controlled trial. Crit Care Med. 1999;27:2346–2350.
Corwin HL, Gettinger A, Pearl RG, et al. Efficacy of recombinant human
erythropoietin in critically ill patients: a randomized controlled trial. JAMA.
Georgopoulos D, Matamis D, Routsis C, et al. Recombinant human erythropoietin therapy in critically ill patients: a dose-response study. Crit
Care. 2005;9:R508–515.
Corwin H, Gettinger A, Fabian T, et al. Efficacy and safety of epoetin alfa
in critically ill patients. N Engl J Med. 2007;357:965.
Singh A, Szczech L, Tang K, et al. Correction of anemia with epoetin alfa
in chronic kidney disease. N Engl J Med. 2006;355:2085.
Wright J, Ung Y, Julian J, et al. Randomized, double-blind, placebocontrolled trial of erythropoietin in non–small-cell lung cancer with diseaserelated anemia. J Clin Oncol. 2007;25:1027–1032.
FDA information for healthcare professionals. Erythropoietin-stimulating
Accessed January 15, 2008.
Sellards AW, Minot GR. Injection of hemoglobin in man and its relation to
blood destruction, with special reference to the anemias. J Med Res. 1916;
Winslow RM. Current status of oxygen carriers (blood substitutes): 2006.
Vox Sang. 2006;91(2):102–110.
Stowell CP. Hemoglobin-based oxygen carriers. Curr Opin Hematol. 2002;
Stowell CP, Levin J, Speiss BD, et al. Progress in the development of RBC
substitutes. Transfusion. 2001;41:287–299.
Moore E. Blood substitutes: the future is now. J Am Coll Surg. 2003;
Scott MG, Kucik DF, Goodnough LT, et al. Blood substitutes: evolution
and future applications. Clin Chem. 1997;43:1724–1731.
Amberson WR, Jennings JJ, Rhode CM. Clinical experience with
hemoglobin-saline solutions. J Appl Physiol. 1949;1:460–489.
Stavitsky JP, Doczi J, Black J, et al. A clinical safety trial with stroma-free
hemoglobin. Clin Pharmacol Ther. 1978;23:73–80.
Intaglietta M, Johnson PC, Winslow RM. Microvascular and tissue oxygen
distribution. Cardiovasc Res. 1996;32:632–643.
Karmaker N, Dhar P. Effect of steady shear stress on fluid filtration through
the rabbit arterial wall in the presence of macromolecules. Clin Exp Pharmacol Physiol. 1996;23:299–304.
Keipert PE, Gonzales A, Gomez CL, et al. Acute changes in systemic
blood pressure and urine output of conscious rats following exchange
transfusion with diaspirin-crosslinked hemoglobin solution. Transfusion.
Przybelski RJ, Dailey EK, Birnbaum ML. The pressor effect of hemoglobin:
good or bad? In: Winslow R, Vandegriff K, Intaglietta M, eds. Advances
in Blood Substitutes: Industrial Opportunities and Medical Challenges.
Boston: Birkhauser;
Hess JR, Macdonald VW, Brinkley WW. Systemic and pulmonary hypertension after resuscitation with cell-free hemoglobin. J Appl Physiol.
Lamy ML, Dailey EK, Brichant JF, et al. Randomized trial of diaspirin
cross-linked hemoglobin solution as an alternative to blood transfusion
after cardiac surgery. Anesthesiology. 2000;92:646–656.
Intaglietta M. Microcirculatory basis for the design of artificial blood. Microcirculation. 1999;6:247–258.
Sprung J, Kindscher JD, Wahr JA, et al. The use of bovine hemoglobin
glutamer-250 (Hemopure) in surgical patients: results of a multicenter, randomized single-blinded trial. Anesth Analg. 2002;94:799–808.
Johnson JL, Moore EE, Offner PJ, et al. Resuscitation with a blood substitutes abrogates pathologic postinjury neutrophil cytotoxic function. J
Trauma. 2001;50:449–456.
Gould S, Moore E, Hoyt D, et al. The first randomized trial of human
polymerized hemoglobin as a blood substitute in acute trauma and emergent
surgery. J Am Coll Surg. 1998;187(2):113–120.
Moore EE, and the PolyHeme Study Group. Postinjury resuscitation with
human polymerized hemoglobin: the U.S.A. multicenter trial. Papers Session, ACS 93rd Clinical Congress, New Orleans, LA, October 10, 2007.
The Bible. New International Version.
Cothren CC, Moore EE, Long JS, et al. Large volume polymerized
haemoglobin solution in a Jehovah’s Witness following abruptio placentae. Transfus Med. 2004;14(3):241–246.
Vaziri K, Roland J, Robinson L, et al. Extreme anemia in an injured Jehovah’s Witness: a test of our understanding of the physiology of severe
anemia and the threshold for blood transfusion. J Trauma Injury Infect
Crit Care In press.