This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citation map  • Citing articles on HighWire  • Citing articles on Web of Science (29)  • Contact me when this article is cited  Related Content  • Similar articles in JAMA  Social Bookmarking   What's this?

Comparison Using Receiver Operating Characteristic Curve Analysis

Jeffrey L. Carson, MD; John M. Eisenberg, MD; Leslie M. Shaw, PhD; Harold L. Kundel, MD; Keith A. Soper, MA

JAMA. 1985;253(5):665-669.

Abstract
We compared the diagnostic characteristics of four assays of prostatic acid phosphatase by two methods. In the first analysis, receiver operating characteristic (ROC) curves were developed and the areas under the curves were calculated. In the second analysis, the sensitivity of each assay was compared with the sensitivities of the other assays with specificities set at equal levels. One or more assays were performed in 1,298 patients; 141 underwent all four assays. Pairwise comparisons of the areas under the ROC curves showed no significant differences. No differences were found in the sensitivities of the four assays when the upper limits of normal were selected to provide equal specificities. We conclude that there is little difference in diagnostic accuracy among the prostatic acid phosphatase assays. Our findings contrast with previous studies that used only one upper limit of normal and that found some assays to be superior to others. Receiver operating characteristic curve analysis corrects for the bias introduced by the choice of an upper limit of normal.

(JAMA 1985;253:665-669)

Author Affiliations
From the Department of Medicine, Section of General Medicine (Dr Eisenberg and Mr Soper), the Department of Pathology and Laboratory Medicine (Dr Shaw), and the Department of Radiology (Dr Kundel), University of Pennsylvania, Philadelphia; and the Department of Medicine, University of Medicine and Dentistry of New Jersey (UMDNJ)/ Rutgers Medical School at Camden (Dr Carson).

Footnotes
Reprint requests to Department of Medicine, UMDNJ/Rutgers Medical School at Camden, Cooper Hospital/University Medical Center, 1 Cooper Plaza, Camden, NJ 08103 (Dr Carson).

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

THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES

Diagnostic performance of the CBCL-Attention Problem Scale as a screening measure in a sample of Brazilian children with ADHD
Lampert et al.
J Atten Disord 2004;8:63-71.
ABSTRACT  

Use of Methodological Standards in Diagnostic Test Research: Getting Better but Still Not Good
Reid et al.
JAMA 1995;274:645-651.
ABSTRACT  

Maximizing Diagnostic Information From the Dexamethasone Suppression Test: An Approach to Criterion Selection Using Receiver Operating Characteristic Analysis
Mossman and Somoza
Arch Gen Psychiatry 1989;46:653-660.
ABSTRACT  

Comparing the Areas under More Than Two Independent ROC Curves
McClish
Med Decis Making 1987;7:149-155.
ABSTRACT  

Decision Analysis: A Progress Report
KASSIRER et al.
ANN INTERN MED 1987;106:275-291.
ABSTRACT