Sensitivity and specificity of a monitoring test.


Journal Article

The usefulness of a diagnostic test is generally assessed by calculating the sensitivity and specificity, or the predictive value positive and predictive value negative of the test. When subjects are monitored periodically for evidence of disease, these calculations must incorporate the varying amounts of information per individual. If in addition, the test results lie on a continuous scale, these quantities vary with the cutoff value (cutpoint) used to define a positive test. They are usually calculated for a spectrum of potential cutpoints in order to produce receiver-operator characteristic curves. In this paper we use a partial likelihood solution to the discrete logistic model in order to obtain estimates of the diagnostic test indices and to provide a significance test when the diagnostic test is administered repeatedly to individuals.

Full Text

Duke Authors

Cited Authors

  • DeLong, ER; Vernon, WB; Bollinger, RR

Published Date

  • December 1985

Published In

Volume / Issue

  • 41 / 4

Start / End Page

  • 947 - 958

PubMed ID

  • 3913467

Pubmed Central ID

  • 3913467

International Standard Serial Number (ISSN)

  • 0006-341X


  • eng

Conference Location

  • United States