Clinical evaluation of free T4 techniques.

This paper describes the specificity and sensitivity performances of the new FT4 immunoextraction methods as compared to FT4I and sensitive TSH measurement. Only the Sclavo Lisophase, Clinical Assays 2 step, Spiria FT4, Mallinckrodt Spac ET and Abbott TDX methods offered improved diagnostic accuracy as compared to FT4I. The use of the albumin dependent analogue FT4 methods decreased test specificity and significantly increased laboratory costs relative to FT4I. Preliminary data is presented concerning the performance of a sensitive TSH assay in the clinical conditions where the diagnostic accuracy of FT4I is compromised. Diagnostic sensitivity for detection of "true" hypo and hyperthyroidism appears excellent as does test specificity when an appropriate reference range is selected. As the different tests had equivalent sensitivity for detection of "true" thyroid disease, a model was developed for the calculation of the relative laboratory cost effectiveness of using each test as a first line test of thyroid function. The model was constructed on the basis of reagent and technical costs, the prevalence of the clinical states, test specificity and the projected clinical and laboratory response to an abnormal result. The model suggested that the use of the albumin-dependent analogue methods would increase costs 3 fold compared to FT4I, whereas the use of the specific FT4 methods (Clinical Assays 2 step, Spiria FT4, Mallinckrodt Spac ET and Abbott TDX) would decrease costs 2 fold compared to FT4I. Sensitive TSH methods were projected to have equivalent or lower costs compared to the specific FT4 methods, depending on whether FT4I or specific FT4 was used as the confirmatory test. It thus appeared that sensitive TSH measurement may become the most cost effective first line test of thyroid function.