PROTEIC PROFILE AND IRON-DEFICIENCY - SIG NIFICANCE OF ALBUMIN-TRANSFERRIN RATIOS

Citation
P. Cacoub et al., PROTEIC PROFILE AND IRON-DEFICIENCY - SIG NIFICANCE OF ALBUMIN-TRANSFERRIN RATIOS, La Revue de medecine interne, 17(8), 1996, pp. 627-634
Citations number
22
Categorie Soggetti
Medicine, General & Internal
ISSN journal
02488663
Volume
17
Issue
8
Year of publication
1996
Pages
627 - 634
Database
ISI
SICI code
0248-8663(1996)17:8<627:PPAI-S>2.0.ZU;2-B
Abstract
From a clinical standpoint the search for iron deficiency is based upo n serum ferritin. However, serum ferritin values may be pathologic in other numerous pathological conditions such as inflammation, liver dis eases, malignant hematologic disorders, hemolysis, etc. Proteic profil e combines the analyze of proteins variations: protein results are con verted in percent of normal values referenced for the technique used. It has been suggested that on the protein profile, an increase in seru m transferrin level compared to a normal serum albumin level (DAT : di fference albumin - transferrin), appears early in the course of iron d eficiency. In order to know the value of a pathologic DAT greater than or equal to 28% in the diagnosis of iron deficiency, we prospectively studied 156 patients consecutively hospitalized in an internal medici ne department. Iron deficiency was defined by a low serum ferritin lev el. Diagnosis performance (sensitivity, specificity, positive and nega tive predictive values) of different biologic markers of iron deficien cy (serum iron, saturation of total iron-binding capacity, low mean er ythrocyte volume) and DAT was compared to the performance of low serum ferritin values. With the exception of low serum ferritin (which have by definition a specificity and a positive predictive value of 100%), pathologic DAT appeared as the best index of iron deficiency with the highest sensitivity (67.4%), specificity (97.3%), positive predicitiv e value (91.2%), negative predicitive value (87.7%) and diagnosis effi cacy (sensitivity x specificity = 0.66). A pathologic DAT associated t o a low serum ferritin level increased the diagnosis performance of bo th tests to 0.72. Diagnosis efficacy of DAT was not changed (0.66) in 83 patients with a confounding factor for ser um ferritin analysis (in flammation, liver diseases, malignant hematologic disorders, hemolysis ) when diagnosis efficacy of all other tests decreased. There was a ne gative correlation between serum ferritin level and DAT level (r = -0. 55; P < 0.0001). In conclusion, an increase of ser urn transferrin of more than 28% compared to serum albumin on a proteic profile gives a s ignificant benefit for the diagnosis of iron deficiency. This benefit increases when data of both DAT and serum ferritin are associated.