Anemia of chronic disease (ACD) is a common form of anemia that occurs in patients suffering from longstanding and/or advanced chronic disease [6]. Patients can be considered to have ACD when they present the following: (1) a chronic infection or inflammation, autoimmune disease or malignancy or renal disease; (2) a hemoglobin concentration <13 g/dl for men and <12 g/dl for women; and (3) a low transferrin saturation (<20 %), but normal or increased serum ferritin concentration (>100 ng/ml) or low serum ferritin concentration (30–100 ng/ml) [7]. Measurement of reticulocyte counts, endogenous erythropoietin (EPO) secretion (ratio of observed EPO to expected EPO), and serum creatinine (glomerular filtration) may be helpful in defining the cause of ACD. Because critically ill patients often have multiple comorbidities, this type of anemia may contribute to the prevalent low hemoglobin levels described on admission to the ICU in large epidemiologic studies [1, 2]. Fifty percent of patients admitted to ICUs with hemoglobin concentrations <10 g/dl have a history of either acute bleeding or ACD [1].

Blood loss is a significant cause of anemia in intensive care patients. Potential sources of blood loss are diagnostic blood sampling and hemorrhage.

Red blood cell (RBC) production, or erythropoiesis, occurs in the bone marrow and is controlled by EPO, a 165 amino acid glycoprotein hormone produced by interstitial fibroblasts in the kidney [12]. EPO promotes the proliferation and differentiation of early erythroid progenitors in the bone marrow into mature erythrocytes. Effective erythropoiesis requires various factors, including iron, zinc, folate and vitamin B₁₂, thyroxine, androgens, cortisol, and catecholamines [13]. RBC formation occurs at a basal rate of 15–20 ml/day under physiological conditions but can increase up to tenfold after hemolysis or heavy blood loss [14].

Critical illness is often associated with increased concentrations of inflammatory cytokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-1, and IL-6, particularly during sepsis. Many of these cytokines have been shown to directly inhibit RBC formation. Other circulating factors, such as interferon-γ, have been shown to induce apoptosis of erythroid precursors in experimental studies. In addition to the relative deficiency of circulating EPO and decreased iron availability, these factors help explain the poor erythroid response to anemia in critically ill patients. Bone marrow hyporeactivity is also suggested by the fact that reticulocyte counts are usually not increased in anemic critically ill patients unless pharmacological doses of EPO are being administered to stimulate erythropoiesis [19, 24].

In healthy humans, erythrocytes have a lifespan of approximately 100–120 days. Normal RBC aging leads to changes in membrane characteristics with decreased deformability, loss of volume and surface area, increased cell density and viscosity, and alterations in the intracellular milieu [13]. These changes result in a decrease in cellular energy levels, increased hemoglobin-oxygen affinity, reduced ability to repair oxidant injury, and decreased ability of the cells to deform when passing through the microvasculature [25]. These changes also indicate that the RBCs are ready for removal by the spleen and reticuloendothelial system. Other determinants of RBC survival include the premature death of mature RBCs (eryptosis) and the removal of RBCs just released from the marrow (neocytolysis). Eryptosis, an apoptosis-like process, is thought to be, in part, triggered by excessive oxidant RBC injury and is inhibited by EPO, which therefore prolongs the lifespan of circulating RBCs [26]. Excessive eryptosis may lead to the development of anemia [27]. Neocytolysis is a process initiated by a sudden decrease in EPO levels by which young circulating RBCs are selectively removed from the circulation [28]. Eryptosis and neocytolysis act at different points in the lifespan of the RBC and thus provide a flexible means of controlling the regulation of total RBC mass.

The normal aging alterations in RBC rheology may occur earlier in critically ill patients, which may have clinical implications [29]. It is likely that critical illness and sepsis, in particular, reduce RBC lifespan, but there is as yet no direct evidence to support this. Experimental data have shown that inflammatory mediators, such as TNF-α and IL-1, can decrease erythrocyte survival time in other settings [30], and oxidative stress has been shown to induce premature apoptosis in RBCs [31].

Hemolysis may be associated with several pathologic conditions, including hemoglobinopathies, hemolytic anemias, bacterial infections, malaria, and trauma. Hemolysis can also occur in conditions in which mechanical forces can lead to RBC rupture, such as surgical procedures, hemodialysis, and blood transfusion. Extracorporeal circuits may lead to complete RBC destruction or cause less severe damage resulting in altered rheological properties. Hemolysis results in release of free plasma hemoglobin and heme, which are toxic to the vascular endothelium [32]. Although most RBC destruction in standard cardiopulmonary bypass procedures can be managed by the endogenous clearing mechanisms, in some cases, for example, in extensive surgery and with prolonged support, higher degrees of hemolysis may occur, and levels of plasma free hemoglobin can rise substantially. These patients are especially susceptible to the toxic influence of un-scavenged RBC constituents and the loss of RBC rheological properties [33].

Hypersplenism may also lead to excessive RBC destruction and is characterized by a significant reduction in one or more of the cellular elements of the blood in the presence of normocellular or hypercellular bone marrow and splenomegaly [34]. In patients with chronic liver disease, hypersplenism secondary to portal hypertension is an important cause of anemia. The main characteristics of hypersplenism are related to the presence of pancytopenia; hemolytic anemia occurs because of intrasplenic destruction of erythrocytes [35].