* This is the Professional Version. *
Evaluation of Anemia
Anemia is not a diagnosis; it is a manifestation of an underlying disorder. Thus, even mild, asymptomatic anemia should be investigated so that the primary problem can be diagnosed and treated.
Acute or chronic blood loss is the first consideration. The diagnosis usually is based on history, examination, and a stool test for occult blood. Further testing for occult bleeding is sometimes necessary.
If blood loss is not detected, laboratory testing is usually done to determine whether anemia is due to deficient RBC production or excessive hemolysis.
The history should address risk factors for particular anemias, symptoms of anemia itself, and symptoms that reflect the underlying disorder.
Anemia has many risk factors. For example, a vegan diet predisposes to vitamin B12 deficiency anemia, whereas alcoholism increases the risk of folate deficiency anemia. A number of hemoglobinopathies are inherited, and certain drugs predispose to hemolysis. Cancer, rheumatic disorders, and chronic inflammatory disorders can suppress bone marrow activity or enlarge the spleen.
The symptoms of anemia are neither sensitive nor specific and do not help differentiate between types of anemias. Symptoms reflect compensatory responses to tissue hypoxia and usually develop when Hb falls to < 7 g/dL. However, they may develop at higher Hb levels in patients with limited cardiopulmonary reserve or in whom the anemia developed very rapidly. Symptoms such as weakness, seeing spots, fatigue, drowsiness, angina, syncope, and dyspnea on exertion can indicate anemia. Vertigo, headache, pulsatile tinnitus, amenorrhea, loss of libido, and GI complaints may also occur. Heart failure or shock can develop in patients with severe tissue hypoxia or hypovolemia.
Certain symptoms may suggest the cause of the anemia. For example, melena, epistaxis, hematochezia, hematemesis, or menorrhagia indicates bleeding. Jaundice and dark urine, in the absence of liver disease, suggest hemolysis. Weight loss may suggest cancer. Diffuse severe bone or chest pain may suggest sickle cell disease, and stocking-glove paresthesias may suggest vitamin B12 or folate deficiency.
Complete physical examination is necessary. Signs of anemia itself are neither sensitive nor specific; however, pallor is common with severe anemia.
Signs of underlying disorders are often more diagnostically accurate than are signs of anemia. Heme-positive stool identifies GI bleeding. Hemorrhagic shock (eg, hypotension, tachycardia, pallor, tachypnea, diaphoresis, confusion—see page Shock) may result from acute bleeding. Jaundice may suggest hemolysis. Splenomegaly may occur with hemolysis, hemoglobinopathy, connective tissue disease, myeloproliferative disorder, infection, or cancer. Peripheral neuropathy suggests vitamin B12 deficiency. Abdominal distention in a patient with blunt trauma suggests acute hemorrhage. Petechiae develop in thrombocytopenia or platelet dysfunction. Fever and heart murmurs suggest infectious endocarditis, a possible cause of hemolysis. Rarely, high-output heart failure develops as a compensatory response to anemia-induced tissue hypoxia.
Laboratory evaluation begins with a CBC, including WBC and platelet counts, RBC indices and morphology (MCV, MCH, MCHC, RBC volume distribution width [RDW]), and examination of the peripheral smear. Reticulocyte count demonstrates how well the bone marrow compensates for the anemia. Subsequent tests are selected on the basis of these results and on the clinical presentation. Recognition of general diagnostic patterns can expedite the diagnosis (see Characteristics of Common Anemias).
Characteristics of Common Anemias
The automated CBC directly measures Hb, RBC count, and MCV (a measure of RBC size). Hct (a measure of the percentage of blood made up of RBCs), MCH (a measure of the Hb content in individual RBCs), and MCHC (a measure of the Hb level in individual RBCs) are calculated values. The diagnostic criterion for anemia in men is Hb < 14 g/dL, Hct < 42%, or RBC < 4.5 million/μL; for women, Hb < 12 g/dL, Hct < 37%, or RBC < 4 million/μL. For infants, normal values vary with age, necessitating use of age-related tables. RBC populations are termed microcytic (small cells) if MCV is < 80 fL, and macrocytic (large cells) if MCV is > 100 fL. However, because reticulocytes are also larger than mature red cells, large numbers of reticulocytes can elevate the MCV and not represent an alteration of RBC production. Automated techniques can also determine the degree of variation in RBC size, expressed as the RDW. A high RDW may be the only indication of simultaneous microcytic and macrocytic disorders (or simultaneous microcytosis and reticulocytosis); such a pattern may result in a normal MCV, which measures only the mean value. The term hypochromia refers to RBC populations in which MCH is < 27 pg/RBC or MCHC is < 30%. RBC populations with normal MCH and MCHC values are normochromic.
The RBC indices can help indicate the mechanism of anemia and narrow the number of possible causes. Microcytic indices occur with altered heme or globin synthesis. The most common causes are iron deficiency, thalassemia, and related Hb-synthesis defects. In some patients with anemia of chronic disease, the MCV is microcytic or borderline microcytic. Macrocytic indices occur with impaired DNA synthesis (eg, due to vitamin B12 or folate deficiencies or chemotherapeutic drugs such as hydroxyurea and antifolate agents) and in alcoholism because of abnormalities of the cell membrane. Acute bleeding may briefly produce macrocytic indices because of the release of large young reticulocytes. Normocytic indices occur in anemias resulting from deficient erythropoietin (EPO) or inadequate response to it (hypoproliferative anemias). Hemorrhage, before iron deficiency develops, usually results in normocytic and normochromic anemia unless the number of large reticulocytes is excessive.
The peripheral smear is highly sensitive for excessive RBC production and hemolysis. It is more accurate than automated technologies for recognition of altered RBC structure, thrombocytopenia, nucleated RBCs, or immature granulocytes and can detect other abnormalities (eg, malaria and other parasites, intracellular RBC or granulocyte inclusions) that can occur despite normal automated blood cell counts. RBC injury may be identified by finding RBC fragments, portions of disrupted cells (schistocytes), or evidence of significant membrane alterations from oval-shaped cells (ovalocytes) or spherocytic cells. Target cells (thin RBCs with a central dot of Hb) are RBCs with insufficient Hb or excess cell membrane (eg, due to hemoglobinopathies or liver disorders). The peripheral smear can also reveal variation in RBC shape (poikilocytosis) and size (anisocytosis).
The reticulocyte count is expressed as the percentage of reticulocytes (normal range, 0.5 to 1.5%) or as the absolute reticulocyte count (normal range, 50,000 to 150,000/μL). Higher values indicate excessive production, or reticulocytosis; in the presence of anemia, reticulocytosis suggests excessive RBC destruction. Low numbers in the presence of anemia indicate decreased RBC production. The reticulocyte response can usually be estimated based on the number of blue-stained cells found when the peripheral smear is stained with a supravital stain; this estimate makes a reticulocyte count, which requires flow cytometry or a large amount of time, unnecessary.
Bone marrow aspiration and biopsy provide direct observation and assessment of RBC precursors. The presence of abnormal maturation (dyspoiesis) of blood cells and the amount, distribution, and cellular pattern of iron content can be assessed. Bone marrow aspiration and biopsy are done to diagnose the following conditions:
Cytogenetic and molecular analyses can be done on aspirate material in hematopoietic or other tumors or in suspected congenital lesions of RBC precursors (eg, Fanconi anemia). Flow cytometry can be done in suspected lymphoproliferative or myeloproliferative states to define the immunophenotype. Bone marrow aspiration and biopsy are not technically difficult and do not pose significant risk of morbidity. These procedures are safe and helpful when hematologic disease is suspected. Both usually can be done as a single procedure. Because biopsy requires adequate bone depth, the sample is usually taken from the posterior (or, less commonly, anterior) iliac crest. If only aspiration is necessary, the sternum may be used.
Serum bilirubin and LDH can sometimes help differentiate between hemolysis and blood loss; both are elevated in hemolysis and normal in blood loss. Other tests are discussed under specific anemias and bleeding disorders (see page Evaluation of Anemia : Testing).
Drug NameSelect Brand Names
chloramphenicolNo US brand name
* This is the Professional Version. *