We compared the clinical value of information on free testosterone as measured with the Coat-A-Count (Diagnostic Products Corp.) radioimmunoassay kit involving a ligand analog with that of total testosterone, the free-androgen index, and free testosterone calculated from concentrations of testosterone, sex-hormone-binding globulin, and albumin, in hirsute women, pregnant women, oral-contraceptive users, women with thyroid disease, and epileptic women taking phenytoin. Total testosterone, the free-androgen index, calculated free testosterone, and free testosterone by RIA were increased in 41-68% of hirsute women. Values for free testosterone increased in the first and third trimesters of pregnancy but remained within normal limits in all non-hirsute groups. Total testosterone was increased in patients having increased sex-hormone-binding globulin, whereas the free-androgen index and, to a lesser extent, calculated free testosterone were significantly decreased. Free testosterone measured by analog RIA not only has greater diagnostic efficiency than total testosterone, it also is technically simpler to determine than the free-androgen index and calculated free testosterone.

We examined the relationship between analytical sensitivity, precision at the lower limit of the reference interval, and diagnostic performance in hyperthyroidism for one radioimmunoassay and five immunometric assay kits for thyrotropin. The analytical sensitivity of these kits extended from 0.05 to 1.56 milli-int. units/L. Diagnostic efficiencies of the immunometric assays, in discriminating between euthyroidism and hyperthyroidism, ranged between 93% and 98%. There was a highly significant correlation (r = 0.99, P less than 0.001) between analytical sensitivity and diagnostic efficiency. The between-assay coefficients of variations, at the lower limit of the reference interval, ranged from 26% to 87%. There was no correlation (r = 0.36) between precision, at this concentration, and diagnostic efficiency. We conclude that analytical sensitivity and not precision is the major determinant in controlling the diagnostic performance of a thyrotropin assay in hyperthyroidism.