A clinical evaluation of a direct radioimmunoassay of testosterone.

We report the evaluation of a commercial coated tube kit (Medgenix) to the measurement of testosterone in plasma/serum samples. The accuracy, precision, specificity and recovery of the assay and the affinity of antibody for ligand were established. The range of concentrations of testosterone providing satisfactory analyses was 0.48-48 nmol/l, which is appropriate for general clinical use. Results of analysis of male and female samples obtained from the routine hospital laboratory in endocrinologically-defined groups were also reviewed. In all groups of male subjects and in normal females, the kit provided results in close agreement to those of an extraction method. With samples from women with various reproductive disorders, as well as with some samples obtained from external quality controls, the kit showed a small but significant negative bias relative to the results of the extraction method. There was, however, no evidence of bias with samples from a separate group of volunteer women with the polycystic ovary syndrome. No interference due to SHBG was detected. Clinical sensitivities using the kit were equivalent to those obtained using the extraction RIA. There was no evidence of spuriously high results in samples from females which has been a common problem with direct radioimmunoassay of testosterone.