Serum testosterone in its total or free form, is a highly valuable diagnostic test and is available in the great majority of clinical laboratories. This reality was possible due to the development of simple and direct assays, adaptable to large automatic systems. Recent publications have called attention to the limitations of these simplified methodologies, mainly in samples with low concentration, as women and children. In this paper we present results obtained using a reference method based on high performance liquid chromatography and tandem mass spectrometry (HPLC/MS-MS) and its comparison with those obtained with a commercial routine immunoassay (electrochemiluminescent assay, ECLIA). Methods were compared in total testosterone measurement (n = 213), as well as in free testosterone evaluation based on calculation inclu-ding sex hormone-binding protein (SHBG) levels (n = 135). Values obtained with ECLIA were significantly higher, with more marked dispersion in low concentration. This phenomenon is clearer when presented as a Bland-Altman plot. Difficulties in the implementation of reference methods as the one presented are discussed, as well as the necessity of caution in the interpretation of values obtained with routine assays, a matter of several publications in recent literature.
Serum total testosterone; Free serum testosterone; Immunoassay; High performance liquid chromatography; Mass spectrometry; Bland-Altman plot
