This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
To the Editor.—
I read with great interest the recent article by Dr Sisson et al (236:1259, 1976) on decision analysis. A direct extension of their evaluation scheme can be used to choose optimally the normal limits for diagnostic tests in which there is overlap between the normal and abnormal values.By knowing how the false-negatives vary with respect to the false-positives for different "normal" limits of a test, one can readily use their method to compute the total number of deaths that would ultimately occur for a variety of "normal" limits. In essence, the ultimate number of deaths can be expressed as a function of the normal limit. That particular limit which minimizes total ultimate mortality would be the optimum normal range.Since the functions involved would be higher order expressions chosen to fit test data, solution would have to be done by readily available numerical techniques on computer.As
Parker WT. Decision Analysis. JAMA. 1977;237(7):641. doi:10.1001/jama.1977.03270340027005
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: