Critical evaluation of fully automated enzyme immunoassays for free thyroxine and thyrotropin.

The technical and diagnostic performance of fully automated immunoassays for free thyroxine and thyrotropin using streptavidin/biotin technology (Enzymun-Test) were examined. The between-assay precision for free thyroxine was 10.4%, 5.4%, 2.5%, 2.3%, 1.1%, and 1.8% at 3.02, 6.27, 17.2, 21.9, 25.6, 42.7 pmol/l; and for thyrotropin was 14.2%, 4.7%, 2.9%, 2.8%, 3.2%, 4.5% at 0.12, 0.46, 1.03, 2.05, 4.8, 12.7 mU/l. The functional detection limit of the assay was 0.09 mU/l. Results for the free thyroxine method correlated well with the IMx (r = 0.91) and the equilibrium dialysis (r = 0.95) assay. Results for the thyrotropin method correlated well with the Tandem-TSH (r = 0.99) and the IMx (r = 0.99) assays. The euthyroid reference range was 11-23 pmol/l and 0.5-3.9 mU/l for free thyroxine and thyrotropin respectively. The free thyroxine assay was not influenced by changes in albumin or thyroxine binding globulin concentration but showed increases at oleic acid concentrations > 4 mmol/l. Spuriously elevated free thyroxine concentration were found in 4 patients, due to assay interference by antibodies in the serum. In a follow up study of 46 patients with non-thyroidal illness, serial measurements showed fluctuating free thyroxine and thyrotropin concentrations with abnormal results occurring in 34%. In a hospital setting, a wider range of free thyroxine (10-28 pmol/l) and thyrotropin (0.22-5.9 mU/l) concentration may be observed in patients who are clinically euthyroid. Abnormal thyroid function tests were however transient and follow up resolved most diagnostic problems.