Van Santen and colleagues recently reported on the significant role of serum hepcidin and erythrocyte (red blood cell) hemoglobin (Hgb) content (RBC-Hgb) in the identification of iron deficiency anemia (IDA) (1); this finding enhances our ability to distinguish IDA from other forms of anemia. In this respect, RBC-Hgb might also be interchangeable with mean corpuscular hemoglobin (MCH), since both parameters quantify the amount of hemoglobin in erythrocytes. Accordingly, if laboratory personnel are unwilling or unable use RBC-Hgb instead of MCH, the latter parameter might prove almost as successful in helping to distinguish IDA from other forms of anemia.
In support of this proposition is the fact that both RBC-Hgb and MCH appear to be superior to mean corpuscular volume (MCV) in establishing the distinction between IDA and other forms of anemia (1, 2). For example, van Santen et al reported that the statistical significance of the difference between IDA and other forms of anemia in terms of the association with the mean RBC-Hgb level was impressive (P < 0.001). In contrast, in terms of association with the mean MCV level, the statistical significance of the difference between IDA and other forms of anemia was modest (P = 0.03). In a study by Francis et al, MCH was the chosen parameter for quantifying the amount of hemoglobin in erythrocytes, and it, too, appeared to be superior to MCV in establishing the distinction between IDA and other forms of anemia (2). Accordingly, in that study, an MCH count of <27 pg was significantly more common than an MCV of <80 fl in patients with IDA (P = 0.016). The van Santen et al and Francis et al studies are comparable because they both enrolled patients with rheumatoid arthritis. Furthermore, to characterize IDA, the 2 studies used serum ferritin cutoff levels that were relatively close to each other, namely, <30 ng/ml versus <20 ng/ml. For these reasons, it would first be useful to ascertain the degree to which RBC-Hgb was correlated with MCH in the patients studied by van Santen et al. Second, it would be of interest to know how the receiver operator characteristic (ROC) analysis of RBC-Hgb compares with the ROC analysis of MCV. Third, when units for measurement of RBC-Hgb are converted into units for measurement of MCH, it would be useful to learn how the results of the ROC analysis for MCH (in the study by van Santen and colleagues) compare with the results of the ROC analysis for MCV.