This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Contact me when this article is cited  Related Content  • Similar articles in JAMA  Social Bookmarking   What's this?

C. E. Miller, MD
Beckman Instruments Inc Brea, Calif

JAMA. 1985;253(23):3396.

	Since this article does not have an abstract, we have provided the first 150 words of the full text PDF and any section headings.

To the Editor.—

Weiss et al,¹ in their article on a screening test for human T-cell leukemia (lymphotropic) retrovirus type III (HTLV-III) antibodies, reported a specificity of 98.6% and a sensitivity of 97.3%, excluding borderline test results. These values of specificity and sensitivity are quite impressive, but they are, in reality, not practical. Borderline results cannot be excluded in actual practice. As the authors point out, if borderline results are considered positive, the specificity and sensitivity are 92.6% and 96.7%, respectively. If the borderline results are considered negative, the specificity and sensitivity are 98.7% and 81.7%, respectively. Excluding the borderline results may give an erroneous impression of the performance of the test.

It would seem that, for the present state of the art, borderline results should be considered positive in screening blood donors. The specificity and sensitivity of a test are applicable only to populations represented by the group . . . [Full Text PDF of this Article]

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?