Bayesian inferences for receiver operating characteristic curves in the absence of a gold standard

Young Ku Choi, Wesley O. Johnson, Michael T. Collins, Ian Gardner

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

Sensitivity and specificity are used to characterize the accuracy of a diagnostic test. Receiver operating characteristic (ROC) analysis can be used more generally to plot the sensitivity versus (1-specificity) overall possible cutoff points. We develop an ROC analysis that can be applied to diagnostic tests with and without a gold standard. Moreover, the method can be applied to multiple correlated diagnostic tests that are used on the same individual. Simulation studies were performed to assess the discrimination ability of the no-gold-standard method compared with the situation where a gold standard exists. We used the area under the ROC curve (AUC) to quantify the diagnostic accuracy of tests and the difference between AUCs to compare their accuracies. In particular, we can estimate the prevalence of disease/infection under the no-gold-standard method. The method we proposed works well in the absence of a gold standard for correlated test data. Correlation affected the width of posterior probability intervals for these differences. The proposed method was used to analyze ELISA test scores for Johne's disease in dairy cattle.

Original languageEnglish (US)
Pages (from-to)210-229
Number of pages20
JournalJournal of Agricultural, Biological, and Environmental Statistics
Volume11
Issue number2
DOIs
StatePublished - Jun 1 2006

Fingerprint

Receiver Operating Characteristic Curve
Bayesian inference
ROC Curve
Gold
gold
Routine Diagnostic Tests
Diagnostic Tests
diagnostic techniques
Operating Characteristics
Specificity
Receiver
Paratuberculosis
Diagnostic Accuracy
Enzyme-linked Immunosorbent Assay
methodology
Dairies
disease prevalence
Score Test
testing
Posterior Probability

Keywords

  • Diagnostic test
  • Markov chain Monte Carlo
  • Sensitivity
  • Serology
  • Specificity

ASJC Scopus subject areas

  • Statistics and Probability
  • Environmental Science(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Agricultural and Biological Sciences(all)
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Bayesian inferences for receiver operating characteristic curves in the absence of a gold standard. / Choi, Young Ku; Johnson, Wesley O.; Collins, Michael T.; Gardner, Ian.

In: Journal of Agricultural, Biological, and Environmental Statistics, Vol. 11, No. 2, 01.06.2006, p. 210-229.

Research output: Contribution to journalArticle

@article{220e54d0162f40178b19fc6d732e717d,
title = "Bayesian inferences for receiver operating characteristic curves in the absence of a gold standard",
abstract = "Sensitivity and specificity are used to characterize the accuracy of a diagnostic test. Receiver operating characteristic (ROC) analysis can be used more generally to plot the sensitivity versus (1-specificity) overall possible cutoff points. We develop an ROC analysis that can be applied to diagnostic tests with and without a gold standard. Moreover, the method can be applied to multiple correlated diagnostic tests that are used on the same individual. Simulation studies were performed to assess the discrimination ability of the no-gold-standard method compared with the situation where a gold standard exists. We used the area under the ROC curve (AUC) to quantify the diagnostic accuracy of tests and the difference between AUCs to compare their accuracies. In particular, we can estimate the prevalence of disease/infection under the no-gold-standard method. The method we proposed works well in the absence of a gold standard for correlated test data. Correlation affected the width of posterior probability intervals for these differences. The proposed method was used to analyze ELISA test scores for Johne's disease in dairy cattle.",
keywords = "Diagnostic test, Markov chain Monte Carlo, Sensitivity, Serology, Specificity",
author = "Choi, {Young Ku} and Johnson, {Wesley O.} and Collins, {Michael T.} and Ian Gardner",
year = "2006",
month = "6",
day = "1",
doi = "10.1198/108571106X110883",
language = "English (US)",
volume = "11",
pages = "210--229",
journal = "Journal of Agricultural, Biological, and Environmental Statistics",
issn = "1085-7117",
publisher = "Springer New York",
number = "2",

}

TY - JOUR

T1 - Bayesian inferences for receiver operating characteristic curves in the absence of a gold standard

AU - Choi, Young Ku

AU - Johnson, Wesley O.

AU - Collins, Michael T.

AU - Gardner, Ian

PY - 2006/6/1

Y1 - 2006/6/1

N2 - Sensitivity and specificity are used to characterize the accuracy of a diagnostic test. Receiver operating characteristic (ROC) analysis can be used more generally to plot the sensitivity versus (1-specificity) overall possible cutoff points. We develop an ROC analysis that can be applied to diagnostic tests with and without a gold standard. Moreover, the method can be applied to multiple correlated diagnostic tests that are used on the same individual. Simulation studies were performed to assess the discrimination ability of the no-gold-standard method compared with the situation where a gold standard exists. We used the area under the ROC curve (AUC) to quantify the diagnostic accuracy of tests and the difference between AUCs to compare their accuracies. In particular, we can estimate the prevalence of disease/infection under the no-gold-standard method. The method we proposed works well in the absence of a gold standard for correlated test data. Correlation affected the width of posterior probability intervals for these differences. The proposed method was used to analyze ELISA test scores for Johne's disease in dairy cattle.

AB - Sensitivity and specificity are used to characterize the accuracy of a diagnostic test. Receiver operating characteristic (ROC) analysis can be used more generally to plot the sensitivity versus (1-specificity) overall possible cutoff points. We develop an ROC analysis that can be applied to diagnostic tests with and without a gold standard. Moreover, the method can be applied to multiple correlated diagnostic tests that are used on the same individual. Simulation studies were performed to assess the discrimination ability of the no-gold-standard method compared with the situation where a gold standard exists. We used the area under the ROC curve (AUC) to quantify the diagnostic accuracy of tests and the difference between AUCs to compare their accuracies. In particular, we can estimate the prevalence of disease/infection under the no-gold-standard method. The method we proposed works well in the absence of a gold standard for correlated test data. Correlation affected the width of posterior probability intervals for these differences. The proposed method was used to analyze ELISA test scores for Johne's disease in dairy cattle.

KW - Diagnostic test

KW - Markov chain Monte Carlo

KW - Sensitivity

KW - Serology

KW - Specificity

UR - http://www.scopus.com/inward/record.url?scp=33744792696&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33744792696&partnerID=8YFLogxK

U2 - 10.1198/108571106X110883

DO - 10.1198/108571106X110883

M3 - Article

AN - SCOPUS:33744792696

VL - 11

SP - 210

EP - 229

JO - Journal of Agricultural, Biological, and Environmental Statistics

JF - Journal of Agricultural, Biological, and Environmental Statistics

SN - 1085-7117

IS - 2

ER -