A Decision Theoretical Comparison of Three Procedures of Screening for a Single Disease
Abstract
We should begin by making a distinction between the terms "screening" and "diagnosis." The purpose of screening is to select from an apparently healthy population those who display sufficient probability of an illness to warrant referral for diagnosis. As defined by the National Conference on Chronic Diseases[47] (cited in [16]): "Screening is the presumptive identification of unrecognized disease or defect by the applications of tests, examinations or other procedures which can be applied rapidly." Thus, screening is not a decision about therapy, but a method of case finding and a step toward diagnosis. Though the emphasis in this paper is on analysis for a single disease, this may be done as a part of a multiple or multiphasic screening program. In fact, the point of interest is how, from a multitude of data, relevant information may be recognized and combined to increase the precision of screening. We shall examine here three ways of using screening data: the single test, with positive or negative indication; the profile, an array of estimates of levels for each of a set of relevant factors; and the index, a single composite of weighted factors.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 07, 1966
- Accession Number
- AD1027658
Entities
People
- Charles D. Flagle
Organizations
- Johns Hopkins University