Analytical evaluation of four sensitive assays of thyrotropin, including effects of variations in patient sampling

Four immunometric kits for thyrotropin with detection limits below 0.1 milli-int. unit/L were evaluated with respect to accuracy, precision, specificity, matrix effects, and high-dose "hook" effect. We also studied variations of values related to the patients' sex and age and time of the day and season when samples were collected. Correlation among the four methods was excellent, except for a few samples, and the interference in these samples could be abolished by adding mouse serum or "suppression medium." These phenomena can also be expected to occur in other immunometric assays involving monoclonal antibodies. With some modifications all tests are suitable for a clinical study of the usefulness of the thyrotropin assay as a primary test for thyroid function.