Anemia and RBC transfusion

Epidemiology

Anemia is defined as a hemoglobin (Hb) level of <130 g/L in men and <120 g/L in nonpregnant women. In the intensive care unit (ICU) population, modest anemia is generally accepted to be a hemoglobin level of ≤100 g/L in both men and women. Large observational studies in all admitted ICU patients have demonstrated that the vast majority will experience an appreciable drop in hemoglobin. In a 2006 European study of over 1000 patients, anemia, as defined by a hemoglobin level of <130 g/L in men or <115 g/L in women, was observed in more than 80% of the patients, and by the time of discharge almost 25% had a hemoglobin level of <90 g/L. Similar observations were made in a 2003 North American critical care study that involved 4892 patients. In this study, 70% of patients developed a hemoglobin level of <120 g/L within 2 days of ICU admission, and 50% had a hemoglobin level of <100 g/L. By the end of the first week at least 97% of patients had a notable decrease in their hemoglobin levels.

Anemia develops early within the first 2 days of ICU admission. In nonbleeding patients, an average hemoglobin drop of 5 g/L/day has been observed, with the largest decline seen in the first days after admission. Persistent anemia is common after ICU discharge, with 77% of patients anemic at hospital discharge and 50% remaining anemic for up to 6 months.

Etiology

The cause of anemia in critical illness is complex and often multifactorial. Primary mechanisms include problems with decreased RBC production, hemodilution from large-volume resuscitation, and increased RBC losses. Production issues include a suppression of bone marrow secondary to a blunted erythropoietin response during critical illness, lack of substrate availability (including iron, vitamin B ₁₂ , and folate), and the presence of renal failure, which leads to an absolute erythropoietin deficiency. ^, The blunted erythropoietin response is in part secondary to the host inflammatory response. This can lead to dysregulation of iron metabolism homeostasis, impaired proliferation of erythroid progenitor cells, and increased inflammatory mediators (interleukin [IL]-1, tumor necrosis factor [TNF]-alpha, increased hepcidin concentrations) causing an iron-restricted anemia. ^, The elevated hepcidin levels lead to iron sequestration in macrophages and reduce iron absorption from the gut, leading to an iron-restricted erythropoiesis.

Anemia caused by blood loss can be subdivided into disease-related and secondary causes. Disease-related causes of anemia include traumatic blood loss, coagulopathy, hemolysis, and gastrointestinal losses. Other secondary losses are primarily iatrogenic and include losses caused by diagnostic sampling, hemolysis, vascular cannulation, renal replacement therapy, and surgical procedures. Daily diagnostic sampling can be an important contributor to anemia. The average volume of blood draws in the ICU results in an average of 40 mL of blood loss per day, and up to 70 mL per day are lost as the patient’s illness progresses. ^, Phlebotomy-related blood loss has been decreased with the introduction of new conservative techniques, with pediatric tubes, reinfusion of the discarded sample from indwelling lines, and reducing the frequency of blood tests.

Consequences of anemia

Anemia is associated with poor outcomes, especially in the elderly and those with chronic disease. ^, The compensatory response of the body in acute or chronic anemia is an extra burden on critically ill patients. The supply/demand balance is particularly stressed in the critically ill anemic patient who has increased cardiac and peripheral oxygen consumption. In patients with ischemic heart disease, coronary flow may be fixed, thereby creating a mismatch between blood supply and oxygen demand. In a large retrospective administrative database study of over 75,000 patients over the age of 65 with myocardial infarction, lower hematocrit (Hct) levels were associated with significantly higher rates of shock and heart failure, in-hospital and 30-day mortality, and increased length of hospital stay.

Red blood cell transfusion

Physiologic effects of RBC transfusion

The negative effects of anemia on oxygen delivery are clear. Although the improved delivery of oxygen after RBC transfusions in these patients has been demonstrated in several studies, an increase in oxygen uptake and consumption by the end organs and tissue beds is less evident and is not a consistent finding. This can be explained by the various adverse factors of stored blood, like low levels of 2,3-diphosphoglycerate (2.3-DPG), with decreases in the ability of hemoglobin to unload oxygen to the tissues, the structural changes in stored blood, and the accumulation of proinflammatory cytokines. Therefore despite a strong physiologic rationale to treat anemia in critically ill patients, particularly those with evidence of end-organ ischemia, studies have failed to demonstrate reliable benefits with respect to oxygen utilization.

Transfusion-related complications

There are infectious and noninfectious risks from RBC transfusions ( Table 125.1 ). RBCs are tested for an extensive number of pathogens, including syphilis, hepatitis B, human immunodeficiency virus (HIV), human T-cell lymphotropic virus, hepatitis C, West Nile virus, Chagas disease, and Zika virus. In addition, storage of blood in refrigerators makes bacterial infection rare. Because of significant enhancements in donor screening and blood testing, the direct transmission of infection through a contaminated blood supply is exceedingly rare (see Table 125.1 ).

Although changes in donor screening and blood testing have led to a significant decline in direct RBC transfusion-related infections over the past 30 years, indirect RBC transfusion-related infections have been the focus of many more recent studies. A systematic review and meta-analysis of 20 randomized trials with 7456 patients evaluated the risk of healthcare–associated infections linked to RBC transfusions and compared different liberal and restrictive transfusion strategies. They found that although a restrictive transfusion strategy was not significantly associated with fewer overall healthcare–associated infections, there was a significant risk reduction in serious infection rates even when controlling for leukoreduction (number needed to treat 48, 95% confidence interval [CI] 36–71). This significance was not observed in a subgroup analysis of 1475 critically ill patients. Another systematic review of randomized trials by the British Medical Journal to study the effect of restrictive versus liberal transfusion strategies used a subgroup meta-analysis of eight trials deemed at lower risk of bias by the authors that included 5107 patients. The authors also showed a lower associated risk of infection with a restrictive transfusion strategy (relative risk [RR] 0.73, 95% CI 0.55–0.98).

Noninfectious complications of RBC transfusions are far more common (see Table 125.1 ). These represent a spectrum from relatively benign (fever) to more severe (acute lung injury) and imminently life-threatening (hemolytic reactions). With increased surveillance and studies focused on complications of transfusions, it has been recognized that transfusion-associated circulatory overload (TACO) is becoming one of the most common risks associated with RBC transfusion. Patients at high risk of volume overload have the greatest danger for developing TACO. The leading cause of RBC transfusion-related morbidity is from transfusion-related acute lung injury (TRALI). This pathophysiologic process is characterized by noncardiogenic pulmonary edema leading to hypoxia. TRALI is secondary to increased lung endothelial permeability and can be caused by immune-mediated and nonimmune-mediated mechanisms. Immune-mediated TRALI is caused by the presence of leukocyte antibodies in the plasma of donor blood that are directed against human leukocyte antigens and human neutrophil antigens. ^, Nonimmune mediation is thought to be the result of biologically active substances such as lipids and cytokines. Any blood product can trigger TRALI, but it is most commonly associated with the transfusion of products containing high plasma content, and the risk increases with the number of products transfused. In the critically ill patient, the clinical picture of TRALI is difficult to distinguish from other causes of acute lung injury. Therefore there has been a proposed update to define TRALI by an expert panel. The authors used the Delphi methodology to split TRALI into two types based on a patient’s risk factor for acute respiratory distress syndrome (ARDS) ( Table 125.2 ). Type I includes those without risk factors for ARDS and is defined similarly to the traditional definition. Type II includes those who may have mild ARDS or risk factors for ARDS, but have had a stable respiratory status for at least 12 hours before their transfusion and otherwise meet the definition of type I.

The exact mechanisms by which these complications occur are not understood, but are likely at least in part attributable to host immune and inflammatory responses. Central to immune-mediated reactions are donor ILs and the TNF, in addition to antibodies or activated neutrophils, fragments of cellular membranes, and soluble human leukocyte antigen, which play important roles in transfusion-induced immunomodulation (TRIM). TRIM is purported to predispose patients to infections and cancer recurrence. Many Western countries across Europe and North America have adopted leukoreduction programs to reduce the effects of TRIM. TRIM may contribute to indirect infectious complications from RBC transfusions, like healthcare–associated infections that occur downstream from the point of transfusion.

Complications from massive transfusions, including coagulopathies, electrolyte disturbances, acid-base imbalances, temperature dysregulation, and citrate toxicity, are also of significant importance but are beyond the scope of this chapter.