Objective: To evaluate prospectively RBC indices as a diagnostic tool at a high complexity general hospital.
Methods: We analyzed 2278 blood cell counts from the core laboratory of our service and we found 343(15%) microcytic anemias. Concomitant serum samples were found from 52 patients above 14 years-old, and ferritin measurement and hemoglobin electrophoresis were performed. We classified our patients in three groups: Iron-deficiency anemia (AF; 26 patients), beta-thalassemia minor (BTM; n=17) and non-iron-deficiency-non-beta-thalassemia (ANFNT; n=9). We evaluated the following RBC indices for group discrimination, with variance and ROC curve analysis: RBC, MCV, MCH, RDW, England's index and Green's index.
Results: None of the indices allowed complete discrimination of all groups. We have established limits for each of the indices, and sensitivity (S), specificity(E), positive and negative predictive values and test efficiency were determined. BTM was better distinguished from the other groups by RBC (above 5 millions/mL) and England's index (<6), both with S=82.3% and E of 82.8%. A RDW above 16% was the best index of AF, with S=69.2% and E= 80.7%.
Conclusions: Presumptive diagnosis of AF is troublesome when using RBC indices; diagnosis should be based on complete laboratory evaluation. Elevated RBC levels, specially above 5 millions/microL, in anemic patients with microcytosis are important indicatives of BTM, and confirmatory tests are suggested.