Allogeneic blood transfusions are given for inadequate oxygen-carrying capacity/delivery and correction of coagulation deficits. Also, blood transfusions provide additional intravascular fluid volume. The American Society of Anesthesiologists (ASA) Committee on Standards and Parameters and a Task Force on Perioperative Blood Management analyzed the literature and solicited many opinions that were published in 2015. The “Practice Guidelines for Perioperative Blood Management” had a major impact on the writing of this chapter, as did the earlier 2006 version of this report, which served as the foundation for this chapter in the 6th edition.
In the past 5 to 10 years, many new conceptual terms have been added to the blood transfusion literature. These terms include transfusion trigger, patient blood management (PBM), transfusion ratios , and preoperative anemia. These terms and concepts have tended to clarify how safety can be enhanced in transfusion medicine. Conversely, a few terms emphasize the severe complications that can occur when multiple transfusions are given to a patient. For example, the term lethal triad describes hypothermia, acidosis, and coagulopathy and is an important negative indicator of transfusion medicine. The 50/50 rule has recently been introduced and has received considerable attention. Basically, a 10% increase in mortality rate was observed with every 10 units of blood given. So, when 50 units of blood are given, there is a 50% mortality rate. While an individual clinician rarely gives 50 units of blood to a patient, the 50/50 rule simply confirms the logical conclusion that patients who require increasing numbers of transfusions have medical or surgical conditions that are very serious with increasing mortality rates. Yet, red blood cell transfusions given for specific clinical situations can decrease mortality rates. Clearly, indications for blood transfusion should be well defined and, if utilized, are often clinically beneficial and even lifesaving.
Refining the preciseness of the indications for blood transfusions is complex. For example, older adult patients are more likely to receive a blood transfusion as compared to younger patients. A recommendation was even made to develop an evidence-based decision aid for blood transfusions.
Blood Therapy Procedures
Determination of the blood types of the recipient and donor is the first step in selecting blood for transfusion therapy. Routine typing of blood is performed to identify the antigens (A, B, Rh) on the membranes of erythrocytes ( Table 24.1 ). Naturally occurring antibodies (anti-B, anti-A) are formed whenever erythrocyte membranes lack A or B antigens (or both). These antibodies are capable of causing rapid intravascular destruction of erythrocytes that contain the corresponding antigens.
| Incidence (%) | ||||
|---|---|---|---|---|
| Blood Group | Antigen on Erythrocyte | Plasma Antibodies | White | African American |
| A | A | Anti-B | 40 | 27 |
| B | B | Anti-A | 11 | 20 |
| AB | AB | None | 4 | 4 |
| O | None | Anti-A, anti-B | 45 | 40 |
| Rh | Rh | 42 | 17 | |
Crossmatch
The major crossmatch occurs when the donor’s erythrocytes are incubated with the recipient’s plasma. Incubation of the donor’s plasma with the recipient’s erythrocytes constitutes a minor crossmatch. Agglutination occurs if either the major or minor crossmatch is incompatible. The major crossmatch also checks for immunoglobulin G antibodies (Kell, Kidd). Type-specific blood means that only the ABO-Rh type has been determined. The chance of a significant hemolytic reaction related to the transfusion of type-specific blood is about 1 in 1000.
Emergency Transfusion
In an emergency situation that requires transfusion before compatibility testing is completed, the most desirable approach is to transfuse type-specific, partially crossmatched blood. The donor erythrocytes are mixed with recipient plasma, centrifuged, and observed for macroscopic agglutination. If the time required to complete this examination (typically < 10 minutes) is not acceptable, the second option is to administer type-specific, non-crossmatched blood if available or else O-negative packed red blood cells. O-negative whole blood is not selected because it may contain high titers of anti-A and anti-B hemolytic antibodies. For adult patients, except female patients of childbearing age, emergency administration of O-positive blood is considered acceptable practice until the patient’s blood type is determined. If the patient’s blood type becomes known and available after 2 units of type O-negative packed red blood cells have been transfused, classic teaching was that subsequent transfusions should probably continue with O-negative blood. However, it is not clear if this practice is necessary and the generally recommended approach is to switch to type-specific blood when it is available.
Soon after blood is typed, crossmatched, and stored, the functional platelets begin to disappear. Fresh whole blood is extremely effective in restoring normal coagulation after severe injury. The effectiveness of fresh whole blood depends on how long it has been stored and its temperature. In the military in Vietnam in the late 1960s, type-specific blood that was maintained at room temperature and stored for no longer than 24 hours was extremely effective in preventing and treating trauma and fluid-induced (e.g., crystalloids) coagulopathies. In the past 50 years, this deduction has been confirmed many times including by retrospective analysis. Not surprisingly, the use of fresh whole blood by forward surgical teams in Afghanistan is associated with improved survival compared to component therapy without platelets.
In an urgent clinical situation, blood needs to be released from the blood bank on an urgent basis. Even a nontrauma hospital should be able to release blood rapidly. At the author’s institution (UCSF Medical Center), a massive transfusion and emergency release protocol ensure that blood products are available at all times. A call activating the massive transfusion protocol will automatically release 4 units of uncrossmatched red blood cells (type O-negative), 4 units of fresh frozen plasma, and 1 unit of platelets. The red blood cells will be released in 5 minutes, with the other products available in 10 minutes. Most acute care hospitals have some type of emergency release or massive transfusion policy.
Type and Screen
Blood that has been typed and screened has been typed for A, B, and Rh antigens and screened for common antibodies. This approach is used when the scheduled surgical procedure is unlikely to require transfusion of blood (hysterectomy, cholecystectomy) but is one in which blood should be available. Blood typing and screening permit more cost-efficient use of stored blood because the blood is available to more than one patient. The chance of a significant hemolytic reaction related to the use of typed and screened blood is approximately 1 in 10,000 units transfused.
Blood Storage
Blood can be stored in a variety of solutions that contain phosphate, dextrose, and possibly adenine at temperatures of 1° C to 6° C. Storage time (70% viability of transfused erythrocytes 24 hours after transfusion) is 21 to 35 days, depending on the storage medium. Adenine increases erythrocyte survival by allowing the cells to resynthesize the adenosine triphosphate needed to fuel metabolic reactions. Changes that occur in blood during storage reflect the length of storage and the type of preservative used. For many years, fresher blood (<5 days of storage) has been recommended for critically ill patients in an effort to improve the delivery of oxygen (2,3-diphosphoglycerate [2,3-DPG] concentrations better maintained). Administration of younger blood (i.e., stored < 14 days) has been associated with better outcomes (e.g., decreased mortality rate and fewer postoperative complications), especially with major surgery. Yet, some authors occasionally conclude that red blood cell quality cannot be determined by duration of storage. More recently, Heddle and associates concluded that the death rate among a general hospital population was not related to the duration of blood storage. Yet, each specialty publishes guidelines for giving blood transfusions, which often includes storage time. These differences even vary within a specialty. Nevertheless, the transfusion-related evidence supported by specialty committees and clinical experience increasingly concludes that the clinician must consider the duration of storage as one of the criteria for selection of a blood product for transfusion.
Decision to Transfuse
The decision to transfuse should be based on a combination of factors: (1) PBM and preoperative anemia; (2) monitoring of blood loss; (3) assessment of how much additional blood loss may occur; (4) monitoring for inadequate perfusion and oxygenation of vital organs; (5) quantitation of intravenous fluids given overall; and (6) monitoring for transfusion indicators, especially the hemoglobin concentration.
Patient Blood Management
PBM has been a major part of our transfusion terminology for the past 5 to 10 years. One of the major components of PBM has been the presence of preoperative anemia. For example, preoperative anemia is a risk factor for a poorer clinical outcome and a predisposing factor for intraoperative blood transfusions. Also, the increasingly common term precision medicine is a broad call for practicing more precise medicine including specifying the indications for a blood transfusion. A major limitation of these conclusions concerns placing attention on one or two variables when many others exist. For example, we cannot forget that hypothermia frequently occurs in patients with severe injury.
Monitoring for Blood Loss
Visual estimation is the simplest technique for quantifying intraoperative blood loss. The estimate is based on a combination of visualization and gravimetric measurements of blood on sponges and drapes and in suction devices. Specifically, differences in weight between dry and blood-soaked gauze pads can routinely be determined. However, these methods for measuring blood loss are only modestly accurate.
Monitoring for Inadequate Perfusion and Oxygenation of Vital Organs
Standard monitors, such as the electrocardiogram and those measuring arterial blood pressure, heart rate, urine output, and oxygen saturation, are commonly used. Analysis of arterial blood gases, mixed venous oxygen saturation, and echocardiography may be useful in selected patients. Tachycardia is an insensitive and nonspecific indicator of hypovolemia, especially in patients receiving a volatile anesthetic. Maintenance of adequate arterial blood pressure and central venous pressure (6 to 12 mm Hg) suggests adequate intravascular blood volume. Urinary output usually decreases during moderate to severe hypovolemia and the resulting tissue hypoperfusion. Arterial pH may decrease only when tissue hypoperfusion becomes severe.
Monitoring for Transfusion Indicators (Especially Hemoglobin Concentration)
The decision to transfuse is based on the risk anemia poses to a patient and the patient’s ability to compensate for decreased oxygen-carrying capacity, as well as the inherent risks associated with transfusion (also see Chapter 20 ). As a member of the UCSF Transfusion Committee for over 20 years, this author can affirm that many of the variables used to guide transfusion therapy are based on clinical judgment rather than peer review studies.
In the past 20 years, new terminology on blood transfusion policy has appeared. A clinician may be using a restrictive blood policy, meaning “give blood only when absolutely necessary.” This restrictive approach evolved many years ago when fear of transmitting hepatitis and human immunodeficiency virus (HIV) was widespread. However, transmission of such diseases is now rare. Blood transfusions given in response to proper indications should decrease patient mortality rates with various conditions. Proper preoperative preparation can reduce the number of blood transfusions used intraoperatively. For example, preoperative anemia should be treated (e.g., with recombinant human erythropoietin and iron). This action decreases not only the need for intraoperative blood transfusions but the overall morbidity and mortality rates.
In parallel with the new terminology, a general standard of care has evolved that healthy patients with hemoglobin values more than 10 g/dL rarely required transfusion, whereas those with hemoglobin values less than 6 g/dL almost always required transfusion, especially when anemia or surgical bleeding (or both) were acute and continuing. Determination of whether intermediate hemoglobin concentrations (6 to 10 g/dL) justify or require transfusion should be based on the patient’s risk for complications of inadequate oxygen delivery. For example, certain clinical situations (e.g., coronary artery disease, chronic lung disease, surgery associated with large blood loss) may warrant transfusion of blood at a higher hemoglobin value than that in otherwise healthy patients. A hemoglobin concentration of 8 g/dL may be an appropriate threshold for transfusion in surgical patients with no risk factors for ischemia, whereas a transfusion threshold of 10 g/dL may be justified in patients who are considered to be at risk for ischemia (emphysema, coronary artery disease). Controlled studies to determine the hemoglobin concentration at which blood transfusion improves outcome in a surgical patient with acute blood loss are few. Yet, to center on hemoglobin values in a complex clinical situation must be done with caution.
More recently, the PBM policy has focused on the words restrictive and liberal for blood transfusions. This policy was dominated by using a hemoglobin value as the indicator. A liberal policy would allow giving blood when hemoglobin levels were more than 9 g/dL. A restrictive policy allowed giving blood only when the hemoglobin levels were preferably 8 g/dL or lower. Analysis of the literature clearly favors the restrictive approach. However, some groups have recommended a liberal approach to sicker patients. One such group is Fominskiy and associates, who wrote, “According to randomized published evidence, perioperative adult patients have an improved survival when receiving a liberal blood transfusion policy.”
Another problem is that the proponents of the restrictive approach do not state what the policy should be for the repetitive or additional administration of blood. Should the indications for the initial administration of blood be the same for each subsequent administration of blood? Clearly the clinician should also estimate whether additional blood will be lost in the actively bleeding patient.
Transfusion of packed red blood cells in patients with hemoglobin concentrations higher than 10 to 12 g/dL does not substantially increase oxygen delivery. Further decreases in the hemoglobin concentration can sometimes be offset by increases in cardiac output. The exact hemoglobin value at which cardiac output increases varies among individuals and is influenced by age, whether the anemia is acute or chronic, and sometimes by anesthesia. For example, the cardiovascular response to anemia in the elderly is decreased, as it is with general anesthesia. Yet, the focus on hemoglobin as a transfusion indicator has existed for many years and still continues. Furthermore, a relatively new noninvasive spectrophotometric monitor (Masimo SpHb) attached to a finger allows the continuous monitoring of hemoglobin levels. Whether this monitor currently can be used for transfusion decisions without a laboratory co-oximeter determination is not clear. For sure, this monitor will provide more opportunity for defining the relationship between hemoglobin levels and transfusion requirements.
The aforementioned considerations indicate that the decision to give a blood transfusion requires a careful thought process that is based on objective clinical indications and a knowledge of transfusion medicine overall.
Blood Components
Packed Red Blood Cells
Packed red blood cells (250- to 300-mL volume with a hematocrit of 70% to 80%) are used for treatment of anemia usually associated with surgical blood loss. The major goal is to increase the oxygen-carrying capacity of blood. Although packed red blood cells can increase intravascular fluid volume, nonblood products, such as crystalloids and colloids, can also achieve that end point. A single unit of packed red blood cells will increase adult hemoglobin concentrations approximately 1.0 to 1.5 g/dL. Administration of packed red blood cells can be facilitated by reconstituting them in crystalloid solutions, such as 50 to 100 mL of saline. The use of hypotonic glucose solutions may theoretically cause hemolysis, whereas the calcium present in lactated Ringer solution may cause clotting if mixed with packed red blood cells.
Complications
Complications associated with packed red blood cells are similar to those of whole blood. An exception would be the chance for development of citrate intoxication, which would be less with packed red blood cells than with whole blood because less citrate is infused. Removal of plasma decreases the concentration of factors I (fibrinogen), V, and VIII as compared with whole blood.
Decision to Administer Packed Red Blood Cells
The decision to administer packed red blood cells should be based on measured blood loss and inadequate oxygen-carrying capacity.
Acute Blood Loss
Acute blood loss in the range of 1500 to 2000 mL (approximately 30% of an adult patient’s blood volume) may exceed the ability of crystalloids to replace blood volume without jeopardizing the oxygen-carrying capacity of the blood. Hypotension and tachycardia are likely, but these compensatory responses may be blunted by anesthesia or other drugs (e.g., β-adrenergic blocking drugs). Compensatory vasoconstriction may conceal the signs of acute blood loss until at least 10% of the blood volume is lost, and healthy patients may lose up to 20% of their blood volume before signs of hypovolemia occur. To ensure an adequate oxygen content in blood, packed red blood cells should be administered when blood loss is sufficiently large. Administration of whole blood, when available, decreases the incidence of hypofibrinogenemia and perhaps coagulopathies associated with administration of packed red blood cells. In the Vietnam conflict, fresh whole blood (typed and crossmatched, but not cooled) was quite effective, especially with massive transfusion-associated coagulopathies. Forty years later in Iraq, military physicians administered fresh whole blood from prescreened “walking donors,” which also can treat or prevent thrombocytopenia. In fact, warm fresh whole blood may be more efficacious than stored component therapy when treating critically ill patients requiring massive blood transfusions. Also, whole blood may be preferable to packed red blood cells when replacing blood losses that exceed 30% of the blood volume. Alternatively, specific ratios of red blood cell transfusions with fresh frozen plasma (FFP) and platelets are being recommended. For example, a ratio of 1.5 units red blood cells with 1 unit of FFP has been proposed. Then 1 unit platelets for 6 units of red blood cells has been recommended in patients with large blood losses and trauma.
With acute blood loss, interstitial fluid and extravascular protein are transferred to the intravascular space, which tends to maintain plasma volume. For this reason, when crystalloid solutions are used to replace blood loss, they should be given in amounts equal to about three times the amount of blood loss, not only to replenish intravascular fluid volume but also to replenish the fluid lost from interstitial spaces. Albumin and hetastarch are examples of solutions that are useful for acute expansion of the intravascular fluid volume. In contrast to crystalloid solutions, albumin and hetastarch are more likely to remain in the intravascular space for prolonged periods (about 12 hours). These solutions avoid complications associated with blood-containing products but do not improve the oxygen-carrying capacity of the blood and, in large volumes (>20 mL/kg), may cause coagulation defects.
Platelets
Administration of platelets allows specific treatment of thrombocytopenia without the infusion of unnecessary blood components. Platelets are derived from volunteer donors (cytapheresis and plateletpheresis). Pooled platelet concentrates are derived from whole blood donation and can be called random-donor platelets . During surgery, platelet transfusions are probably not required unless the platelet count is less than 50,000 cells/mm 3 as determined by laboratory analysis or in predetermined ratios with red blood cells as described previously.
Complications
The risks associated with platelet concentrate infusions are (1) sensitization reactions because of human leukocyte antigens on the cell membranes of platelets and (2) transmission of infectious diseases, which is rare. One of the leading causes of transfusion-related fatalities in the United States is bacterial contamination, which is most likely to occur in platelet concentrates ( Table 24.2 ). Platelet-related sepsis can be fatal and occurs as frequently as 1 in 5000 transfusions; it is probably underrecognized because of the many other confounding variables present in critically ill patients. When donor platelets are cultured before infusion (and not released until the culture is negative after a minimum of 24 hours’ incubation), the incidence of sepsis may be significantly reduced, but sepsis is still possible. The fact that platelets are stored at 20° C to 24° C instead of 4° C probably accounts for the greater risk of bacterial growth than with other blood products. As a result, any patient in whom a fever develops within 6 hours of receiving platelet concentrates should be considered to possibly be manifesting platelet-induced sepsis, and empirical antibiotic therapy should be instituted.
