Evaluation of Anemia in Adults: Going after the Common Causes.

Unexplained anemia is a common problem in adult medicine. Traditional approaches to etiologic diagnosis have included testing for “nutritional” deficiencies such as serum vitamin B12, folic acid, iron and iron binding capacity (IBC)(often a house staff approach); performing upper and lower endoscopy (often the Gastroenterologist or the Surgeon); or review of peripheral blood and bone marrow smears and biopsy (the Hematologist). Decision trees starting from MCVs and reticulocyte counts are commonly suggested in Hematology teaching manuals. However none of these approaches are based on knowledge of the most likely causes of anemia in the population to be tested. We hoped to improve on these largely unfounded and unevaluated empiric strategies by determining the actual frequencies of different causes of anemia among non-pregnant adults (20 and older) with initially unexplained anemias (Hb <12 in men, <11 in women) in our hospital, and have done such studies on three occasions in the last 15 years at Yale New Haven Hospital. Patients selected were 1) 202 consecutive adults with anemia new to the hospital laboratory database; 2) 800 consecutive adults treated, or retreated, with iron in the hospital, focussing the analysis on the 200 who were anemic and not just given iron because they were post-operative or post-partum; and 3) 100 consecutive adults having what we would consider “anemia tests” drawn for work up (such as reticulocyte counts, or vitamin B12 assays, etc). Using standard criteria for diagnosis of different causes of anemia, and, mostly, available data, we were able to classify 80–90% of the anemic patients in each study population. From each of these patient groups we were able to draw pie charts showing relative frequencies of different causes of anemia. We recognized a total of 14 causes and groups of causes of anemia (we grouped hemoglobinopathies as one category, for example) in our hospital population. But the four commonest causes were the same in all three studies: anemia of chronic (inflammatory) disease (ACD) (24, 30, and 37%), acute, missed G I bleed (17, 16, and 13%), iron deficiency anemia (IDA)(13, 23, and 16%) and anemia of chronic renal insufficiency (13, 28, and 14%). Popular targets of work up such as B12 and folate deficiency, and hemolytic anemias, together constituted less than 4% of each study population. We therefore propose as initial work up/evaluation of unexplained anemia in adults the following tests: review of the patient’s history for ACD causes; erythocyte sedimentation rate; iron and IBC (and hepcidin when available); stool exam for blood, especially non-occult blood and melena, before considering colonoscopy; ferritin; attention to the probably already available MCV, RDW, and creatinine; and serum erythropoietin (relating its result to Hb and Hct). According to our hospital epidemiologic studies this approach should reveal the causes of anemia in 67–97% of adults. Clues to other specific anemias learned from the patient, such as a history of sickle cell disease, or the technologist’s discovery of many spherocytes on the blood smear should of course be attended to, as priorities. And, although helpful in only two of the four commonest anemias (ACD, IDA), review of a good blood smear is an important cross-check that occasionally yields surprises (Please see BJ Bain, NEJM, 8/4/05).