Modern differential diagnosis of various types of anemia in patients with HIV-associated disease

This article substantiates the development of a diagnostic algorithm to differentiate between anemia of chronic disease (ACD) and iron deficiency anemia (IDA) in patients with HIV-associated disease. A cross-sectional descriptive study included 125 patients with HIV-associated disease, 101 of whom had anemia syndrome and 24 were without it (control group). Patients with anemia were divided into three groups: Group 1 with ACD only, Group 2 with ACD and IDA, and Group 3 with IDA only. Upon admission, all patients underwent assessments for red blood cell count, hemoglobin, IL-6, IL-10, and IL-1β, IFN-γ, TNF-α, ferritin, C-reactive protein, transferrin, hepcidin, and soluble transferrin receptor levels. Discriminant analysis followed by obtaining a canonical linear discriminant function was used to calculate a mathematical model for the differential diagnosis of ACD and IDA. Canonical analysis was also used to obtain centroids for ACD (2.86 arbitrary units) and IDA (2.54 arbitrary units). A patient with a calculated canonical linear discriminant function was to be assigned to the group of ACD or IDA based on the minimal distance to the corresponding centroid. The resulting mathematical model has 88.8% sensitivity and 100% specificity. The use of IL-6 and ferritin concentrations is also proposed as independent laboratory markers for the differential diagnosis of ACD and IDA. The canonical linear discriminant function obtained from the calculations, along with the levels of IL-6 and ferritin in the blood, have high diagnostic value for verifying ACD and IDA in patients with HIV-related disease. Their use enables the determination of the type of anemia during the initial contact between the physician and the patient.