1799

Chapter 47 - Transfusion Therapy

Ronald D. Miller

BLOOD THERAPY

According to a survey conducted by the Committee on Blood and Blood Products of the American Society of Anesthesiologists,^[1] much of all blood given to patients is during the perioperative period. The anesthesiologist should be an expert on the implications and the complications associated with blood transfusions and should be a leader of acute transfusion medicine in the hospital setting.

Indications for Transfusion

Allogeneic (Homologous) Blood

Blood transfusions are given to increase oxygen-carrying capacity and intravascular volume. Theoretically, increasing vascular volume is not an indication for blood transfusions, because volume can be augmented with administration of fluids that do not transmit infections (e.g., crystalloids or some colloids). On the practical side, when a patient is hemorrhaging, blood is given to increase oxygen-carrying capacity and intravascular volume. However, increasing oxygen-carrying capacity is the only real indication for blood transfusions.

What are the specific indications for blood transfusions? With a specific indication (i.e., increase oxygen-carrying capacity), this question, at first glance, should be easy to answer. Unfortunately, there is little or no way, except in severe life-and-death situations, to determine whether oxygen-carrying capacity is inadequate. The controversy revolves around how much emphasis should be placed on hemoglobin or hematocrit values. The limitation of such values is based on a potential extreme variability from one patient to another regarding their need for increased oxygen-carrying capacity. For example, young healthy patients with normal cardiorespiratory function may easily compensate for anemia (i.e., chronic or acutely induced by hemorrhage), whereas at an identical hematocrit value, elderly patients with cardiac disease may have serious problems with surgery and anesthesia. An individual patient's hemoglobin level may vary markedly in the perioperative period independent of and in addition to transfusions of red blood cells (RBCs). This concern has led many medical organizations to place emphasis on overall medical judgment rather than a specific laboratory value,^{[2] [3]} with some debate.^[4] Conversely, other organizations have chosen to make far more complex directions, such as those in Table 47-1 from the American College of Surgeons. Despite the intellectual appeal of the latter approach, most of the American College of Surgeons' criteria cannot distinguish inadequate intravascular volume from diminished oxygen-carrying capacity. I think the hemoglobin concentration should provide the basis for determining the transfusion requirements for each patient.

The foundation for using the hemoglobin or hematocrit value as the initial consideration for defining transfusion requirements followed a 1988 National Institutes of Health

1800

The ultimate determination of the hematocrit or hemoglobin value at which blood should be given will have to be a clinical judgment based on many factors, such as cardiovascular status, age, anticipated additional blood loss, arterial oxygenation, mixed venous oxygen tension, cardiac output, and blood volume. Oxygen extraction ratio has been recommended as an indicator for transfusions,^[6] but this technique requires invasive monitoring. Even so, the results by this indicator were not dramatic between groups who were or were not transfused. To further complicate this issue, indications for blood transfusions also probably depend on the source of the blood. For example, indications for autologous blood may be more liberal because it is unlikely to transmit diseases (e.g., hepatitis, acquired immunodeficiency syndrome [AIDS]) than is allogeneic blood. However, autologous blood should not be viewed as completely safe because of the possibility of laboratory error or a hemolytic transfusion reaction.^[7] Even febrile nonhemolytic and allergic reactions can occur after administration of autologous blood^[7] (see Chapter 48 ).

Several groups have attempted to define, usually in intensive care patients, the point at which blood transfusions should be given by measures of tissue oxygenation and hemodynamics (e.g., increase in oxygen consumption in response to added oxygen content).^{[8] [9] [10] [11] [12] [13]} No specific measure could consistently predict when a patient would benefit from a blood transfusion. However, evidence suggests that the quality (e.g., age) of the blood and its increased oxygen capacity (e.g., hemoglobin level >10 g/dL) may benefit very sick patients. One study found that when blood stored for more than 15 days was given, evidence of splanchnic ischemia occurred.^[8] This concept was confirmed later by Purdy and colleagues,^[14] who found that patients who received 17-day-old blood (range, 5 to 35 days) versus 25-day-old blood (range, 9 to 36 days) had a higher survival rate. In 1999, the length of storage of banked blood was associated with the development of postoperative pneumonia after cardiac surgery.^[15] However, in 2003, prolonged storage of blood was not associated with increased morbidity after cardiac surgery.^[16] Because the quality of blood decreases with length of storage, an association with morbidity would not be surprising. The influence of the age of blood infused is discussed later in this chapter.

Perhaps the development of more sensitive indicators of tissue oxygenation (e.g., intramucosal pH) will provide indicators for transfusion. Using outcome data for an orthopedic surgery population, variations in hemoglobin levels were found to be unrelated to duration of hospitalization,^[17] but trained athletes and postoperative cardiac patients showed improved physical capabilities when hemoglobin levels were increased.^[18] Conversely, Weiskopf and coworkers^[19] found that decreases in hemoglobin concentration to 5.0 g/dL did not produce any evidence of inadequate oxygenation in healthy patients. However, these patients were not subjected to the stresses of recovery from surgery and anesthesia. Weiskopf and associates ^[19] found that these patients compensated for their low hemoglobin levels by increasing heart rate and stroke volume. It is tempting to argue that patients who have a more rapid than expected heart rate or who cannot increase their cardiac output should receive a transfusion to a higher hemoglobin level than 10 g/dL. Unfortunately, precise conclusions cannot be derived from these helpful and suggestive data.

To arrive at some conclusions in the presence of incomplete data, two complementary recommendations are given. In 1996, the American Society of Anesthesiologists Practice Guidelines^[2] offered these recommendations:

1. Transfusion is rarely indicated when the hemoglobin concentration is greater than 10 g/dL and is almost always indicated when it is less than 6 g/dL, especially when the anemia is acute.
2. The determination of whether intermediate hemoglobin concentrations (6 to 10 g/dL) justify or require RBC transfusion should be based on the patient's risk for complications of inadequate oxygenation.

---

1801

3. The use of a single hemoglobin "trigger" for all patients and other approaches that fail to consider all important physiologic and surgical factors affecting oxygenation is not recommended.
4. When appropriate, preoperative autologous blood donation, intraoperative and postoperative blood recovery, acute normovolemic hemodilution, and measures to decrease blood loss (i.e., deliberate hypotension and pharmacologic agents) may be beneficial.
5. The indications for transfusion of autologous RBCs may be more liberal than those for allogeneic RBCs because of less frequent (but still significant) risks associated with the former.

With the help of Habibi and colleagues^[20] in 1998, the following indications were recommended, with the rule of thumb that administration of 1 unit of packed red blood cells (PRBCs) will increase hematocrit value by 3% to 5%:

1. Blood loss greater than 20% of blood volume when more than 100 mL
2. Hemoglobin level less than 8 g/dL
3. Hemoglobin level less than 10 g/dL with major disease (e.g., emphysema, ischemic heart disease)
4. Hemoglobin level of less than 10 g/dL with autologous blood
5. Hemoglobin level less than 12 g/dL and ventilator dependent

Although the recommendations of Habibi and coworkers^[20] are current, the elusive "transfusion trigger" remains in a prominent part of the debates in anesthesia specifically and in medicine in general. Both recommendation lists^{[2] [20]} agree that a transfusion trigger of 8.0 g/dL or less can be tolerated by patients who are not critically ill or do not have severe cardiorespiratory disease. As of 2004, that conclusion is still valid.

In the past 5 years, research has been directed toward how liberal the transfusion trigger should be in critically ill patients. In 1999, Herbert and associates^[21] found that in critically ill patients, a restrictive transfusion strategy (hemoglobin level < 10gm/dL) was as effective as a more liberal transfusion (hemoglobin level = 10 to 12 gm/dL) except in patients with acute myocardial infarction and unstable angina. Despite the problem with identifying a specific transfusion trigger, Ely and Bernard^[22] have generally confirmed the conclusions of Habibi and colleagues. Subsequent studies have found no apparent benefit or a trend toward better outcomes with transfusion-induced hemoglobin levels more than 8.0 g/dL (i.e., 9.0 to 10.0 g/dL).^{[23] [24] [25] [26]} Subsequent efforts and editorials have leaned toward a lower transfusion trigger for even critically ill patients.^{[27] [28] [29]} Perhaps the margin of safety should be increased by increasing the hemoglobin in critically ill patients, including those with cardiorespiratory disease. There is increasing concern about the underuse of blood transfusion therapy. ^[30] However, blood is a valuable resource that sometimes is in short supply. Perhaps with the availability of erythropoietin and synthetic RBCs, a more liberal hemoglobin concentration can be used. As concluded by Weiskopf,^[31] "We merely await advances in technology that will enable us to measure directly the value of concern and thereby free us from arguments over which surrogate (e.g., hemoglobin) to measure and what value indicates the need for augmented oxygen delivery."

Autologous Blood

Autologous blood is assumed to be (and probably is) much safer than allogeneic blood, mainly because of the decreased risk of infection. Because of a marked decrease in infectivity from allogeneic blood (see "Infectivity of Blood"), the difference in safety compared with autologous blood is much less.^{[32] [33]} For example, Kanter and coworkers^[34] found that 25 of 140 patients undergoing hysterectomy who donated blood received blood transfusions, whereas 1 of 123 patients who did not donate autologous blood was transfused. They concluded that elimination of preoperative autologous blood donation does not increase the risk of allogeneic blood transfusion. They further argued that autologous blood does have risks. One of every 16,000 autologous blood donations results in a reaction severe enough to require hospitalization.^[35] Complications associated with autologous blood transfusions include the following:

1. Anemia
2. Preoperative myocardial ischemia from anemia
3. Administration of the wrong unit (1:100,000)
4. Need for more frequent blood transfusions
5. Febrile and allergic reactions^[7]

However, the relative value of autologous blood is based on the assumption that the risks of allogeneic blood have been properly assessed. The advocates for the elimination of autologous blood transfusion have not included immunosuppression or outbreaks of other infections in allogeneic blood. For example, from 1986 to 1991, 182 transfusion-associated fatalities were reported to the U.S. Food and Drug Administration (FDA), 29 (16%) of which were caused by bacterial contamination.^[36] Since then, 10 cases of Yersinia enterocolitica have been reported. All cases involved allogeneic blood, although two patients also received autologous blood. Autologous blood can affect the immune system (see "Transfusion-Related Immunomodulation"). The testing and screening of blood donors is by no means perfect. ^[38] Ask yourself a question: If given a choice, would you want your own blood or allogeneic blood?

Nevertheless, the conclusion still is that autologous blood is preferred over allogeneic (homologous) blood with a more generous hemoglobin level (<10 g/dL) than that at which autologous blood can be given.^[38]