Original article
Evaluation of diagnostic tests with multiple diagnostic categories

https://doi.org/10.1016/0895-4356(88)90051-0Get rights and content

Abstract

The evaluation of diagnostic tests attempts to obtain one or more statistical parameters which can indicate the intrinsic diagnostic utility of a test. Sensitivity, specificity and predictive value are not appropriate for this use. The likelihood ratio has been proposed as a useful measure when using a test to diagnose one of two disease states (e.g. disease present or absent). In this paper, we generalize the likelihood ratio concept to a situation in which the goal is to diagnose one of several non-overlapping disease states. A formula is derived to determine the post-test probability of a specific disease state. The post-test odds are shown to be related to the pre-test odds of a disease and to the usual likelihood ratios derived from considering the diagnosis between the target diagnosis and each alternate in turn. Hence, likelihood ratios derived from comparing pairs of diseases can be used to determine test utility in a multiple disease diagnostic situation.

References (6)

  • R.H. Fletcher et al.

    Clinical Epidemiology-The Essentials

    (1982)
  • C.H. Goldsmith

    Generalized measures of sensitivity, specificity and predictive value in epidemiologic studies

    Am J Epidemiol

    (1977)
  • R.J. Baldessarini et al.

    The predictive power of diagnostic tests and the effect of prevalence of illness

    Arch Gen Psychiat

    (1983)
There are more references available in the full text version of this article.

Cited by (15)

  • Real world-like simulations show efficient predictive power of in vitro skin corrosion tests used as stand-alone and in combination and how can toxicologists take advantage of them

    2021, Toxicology in Vitro
    Citation Excerpt :

    For the SCTs of this paper and before that the test T is performed, the odds of finding an actual positive in the whole population of chemicals is the probability of being of an actual positive which is the Prevalence Pe divided by the probability of not being an actual positive (1-Pe) i.e., Prevalence ÷ (1-Prevalence) = Pe/(1-Pe). This odds is called Pre-Test Odds (i.e. Pre-Test Odds = Pe/(1-Pe)) (Birkett, 1988; Altman and Bland, 1994; Deeks and Altman, 2004). For data under consideration in this paper:

  • The indication area of a diagnostic test. Part i - Discounting gain and loss in diagnostic certainty

    2015, Journal of Clinical Epidemiology
    Citation Excerpt :

    In part II of this study, we demonstrate the consequences of attaching different utilities to the test results for the indication area and diagnostic max [33]. We developed this approach in its most simple form for a binary diagnostic test result (positive or negative) and for a binary disease outcome (diseased or healthy) [41]. The model can be expanded to more complex testing situations, for example, if one has to choose between several tests simultaneously or tests in sequential order.

  • Incorporating severity of illness into estimates of likelihood ratios

    1990, American Journal of the Medical Sciences
View all citing articles on Scopus
View full text