To the Editor.
—When mild thyroid dysfunction is to be detected, much depends on the criterion used for final diagnosis. In the article by dos Remedios et al published in the Archives (1980;140:1045-1049), "the final diagnosis was made clinically by physicians experienced in thyroid disease." If it were possible to come anywhere near 100% accuracy in such a way, laboratory tests would be of little use. (The parameter to be evaluated should, of course, not be included.) My doubts are strengthened by their finding that the free thyroxine index (FT4I) had 100% sensitivity, an opinion adopted in an editorial by Martin Nusynowitz, MD, in the same issue (p 1017). It is, however, well known that not only "subclinical" but also mild clinical hypothyroidism (responding to levothyroxine sodium therapy) can occur in the presence of a normal FT4I; serum thyroid-stimulating hormone (TSH) levels used to be elevated