Sage Journals: Discover world-class research

Abstract

A receiver operating characteristic curve (ROC) is a statistical tool to assess the accuracy of a test. In this article, we propose to review the methods and applications of the ROC curve for various situations in clinical trials and drug development. The ROC curve is widely accepted as a method for selecting an optimal cutoff point and for comparing the accuracy of diagnostic tests. It is a plot of the true positive rate against the false positive rate for different cutoff points. The sensitivity and specificity of a test, however, depend on the level that has been chosen as a cutoff point for normal or abnormal. In bioavailability or bioequivalence (BA/BE) studies, the area under the curve (AUC) is often used to measure the extent of absorption or total amount of drug absorbed into the body. AUC is usually calculated using a trapezoidal rule based on blood or plasma concentrations obtained at various sampling time points. In pharmacovigilance, the ROC curve is used in signal detection.

Keywords

ROC curve Clinical trial Signal detection Sensitivity Specificity AUC

Get full access to this article

View all access options for this article.

References

Altman

Bland

Diagnostic tests 3: receiver-operating characteristic plots. BMJ. 1994;309:188.

Akobeng

. Understanding diagnostic tests 3: receiver operating characteristic curves. Acta Pædiatrica. 2007;96: 644–647.

Pickard

. Using receiver operating characteristic (ROC) curves to evaluate digital mammography. University of Texas Health Science Center.

Pepe

. The Statistical Evaluation of Medical Tests for Classification and Prediction. New York: Oxford University Press; 2003.

Zhou

X-H

Obuchowsk

McClish

. Statistical Methods in Diagnostic Medicine. New York: Wiley; 2002.

Zweig

Campbell

Receiver operating characteristic (ROC) curve: a fundamental evaluation of clinical medicine. Clin Chem. 1993;39: 561–577.

Goldstein

Mushlin

. Use of a single throxine test to evaluate ambulatory medical patients for suspected hypothyroidism. J Intern Med. 1987;2: 20–24.

van Puijenbroek

. Quantitative signal detection in pharmacovigilance. PhD diss., Utrecht University, 2001.

van Puijenbroek

Bate

Leufkens

A comparison of measures of disproportionality for signal detection in spontaneous reporting systems for adverse drug reactions. Pharmacoepidemiol Drug Saf. 2002;11: 3–10.

10.

Stricker

Saty

BMP

. Detection, verification, and quantification of adverse drug reactions. BMJ. 2004;329: 44–47.

11.

Maxion

Roberts

. The proper use of ROC curves in intrusion/anomaly detection. Technical Report Series CS-TR-871. University of Newcastle Upon Tyne; 2004.

12.

Mann

Andrews

. Pharmacovigilance, 2nd ed. London: Wiley; 2002.

13.

Eudravigilance Expert Working Group. Guideline on the use of statistical signal detection methods in the Eudravigilance data analysis system. London: European Medicines Agency; November 16, 2006.

14.

Lindquist

Data quality management in pharmacovigilance. Drug Saf. 2004;27: 857–870.

15.

Patterson

Jones

Bioequivalence and Statistics in Clinical Pharmacology. New York: Chapman & Hall/CRC; 2006.

16.

Yeh

. Using Trapezoidal Rule for the Area Under a Curve Calculation. Cary, NC: SAS Institute.

17.

Rick

. Drugs From Discovery to Approval. New York: Wiley; 2004.

18.

Gibaldi

Perrier

Pharmacokinetics, 2nd ed. New York: Informa Healthcare; 2007.

19.

Dipiro

Spruill

Wade

Blouin

Pruemer

. Concepts in Clinical Pharmacokinetics, 4th ed. New York: ASHP; 2005.

20.

Shargel

Wu-Pong

ABC

. Applied Biopharmaceutics and Pharmacokinetics, 5th ed. New York: McGraw-Hill; 2005.

A Review of Methods and Applications of the ROC Curve in Clinical Trials

Abstract

Keywords

Get full access to this article

References