Modelling risk when binary outcomes are subject to error

Pat McInturff, Wesley O. Johnson, David Cowling, Ian Gardner

Research output: Contribution to journalArticle

80 Citations (Scopus)

Abstract

We present methods for binomial regression when the outcome is determined using the results of a single diagnostic test with imperfect sensitivity and specificity. We present our model, illustrate it with the analysis of real data, and provide an example of WinBUGS program code for performing such an analysis. Conditional means priors are used in order to allow for inclusion of prior data and expert opinion in the estimation of odds ratios, probabilities, risk ratios, risk differences, and diagnostic test sensitivity and specificity. A simple method of obtaining Bayes factors for link selection is presented. Methods are illustrated and compared with Bayesian ordinary binary regression using data from a study of the effectiveness of a smoking cessation program among pregnant women. Regression coefficient estimates are shown to change noticeably when expert prior knowledge and imperfect sensitivity and specificity are incorporated into the model.

Original languageEnglish (US)
Pages (from-to)1095-1109
Number of pages15
JournalStatistics in Medicine
Volume23
Issue number7
DOIs
StatePublished - Apr 15 2004

Fingerprint

Binary Outcomes
Specificity
Diagnostic Tests
Routine Diagnostic Tests
Sensitivity and Specificity
Imperfect
Odds Ratio
Modeling
Risk Difference
WinBUGS
Binary Regression
Coefficient Estimates
Expert Opinion
Regression Estimate
Bayes Factor
Smoking
Expert Testimony
Smoking Cessation
Regression Coefficient
Prior Knowledge

Keywords

  • Bayes factor
  • Bayesian
  • Logistic regression
  • Markov chain Monte Carlo
  • Sensitivity
  • Specificity

ASJC Scopus subject areas

  • Epidemiology

Cite this

Modelling risk when binary outcomes are subject to error. / McInturff, Pat; Johnson, Wesley O.; Cowling, David; Gardner, Ian.

In: Statistics in Medicine, Vol. 23, No. 7, 15.04.2004, p. 1095-1109.

Research output: Contribution to journalArticle

McInturff, Pat ; Johnson, Wesley O. ; Cowling, David ; Gardner, Ian. / Modelling risk when binary outcomes are subject to error. In: Statistics in Medicine. 2004 ; Vol. 23, No. 7. pp. 1095-1109.
@article{2badf2789a754816a4da5a8e8bc9238a,
title = "Modelling risk when binary outcomes are subject to error",
abstract = "We present methods for binomial regression when the outcome is determined using the results of a single diagnostic test with imperfect sensitivity and specificity. We present our model, illustrate it with the analysis of real data, and provide an example of WinBUGS program code for performing such an analysis. Conditional means priors are used in order to allow for inclusion of prior data and expert opinion in the estimation of odds ratios, probabilities, risk ratios, risk differences, and diagnostic test sensitivity and specificity. A simple method of obtaining Bayes factors for link selection is presented. Methods are illustrated and compared with Bayesian ordinary binary regression using data from a study of the effectiveness of a smoking cessation program among pregnant women. Regression coefficient estimates are shown to change noticeably when expert prior knowledge and imperfect sensitivity and specificity are incorporated into the model.",
keywords = "Bayes factor, Bayesian, Logistic regression, Markov chain Monte Carlo, Sensitivity, Specificity",
author = "Pat McInturff and Johnson, {Wesley O.} and David Cowling and Ian Gardner",
year = "2004",
month = "4",
day = "15",
doi = "10.1002/sim.1656",
language = "English (US)",
volume = "23",
pages = "1095--1109",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "7",

}

TY - JOUR

T1 - Modelling risk when binary outcomes are subject to error

AU - McInturff, Pat

AU - Johnson, Wesley O.

AU - Cowling, David

AU - Gardner, Ian

PY - 2004/4/15

Y1 - 2004/4/15

N2 - We present methods for binomial regression when the outcome is determined using the results of a single diagnostic test with imperfect sensitivity and specificity. We present our model, illustrate it with the analysis of real data, and provide an example of WinBUGS program code for performing such an analysis. Conditional means priors are used in order to allow for inclusion of prior data and expert opinion in the estimation of odds ratios, probabilities, risk ratios, risk differences, and diagnostic test sensitivity and specificity. A simple method of obtaining Bayes factors for link selection is presented. Methods are illustrated and compared with Bayesian ordinary binary regression using data from a study of the effectiveness of a smoking cessation program among pregnant women. Regression coefficient estimates are shown to change noticeably when expert prior knowledge and imperfect sensitivity and specificity are incorporated into the model.

AB - We present methods for binomial regression when the outcome is determined using the results of a single diagnostic test with imperfect sensitivity and specificity. We present our model, illustrate it with the analysis of real data, and provide an example of WinBUGS program code for performing such an analysis. Conditional means priors are used in order to allow for inclusion of prior data and expert opinion in the estimation of odds ratios, probabilities, risk ratios, risk differences, and diagnostic test sensitivity and specificity. A simple method of obtaining Bayes factors for link selection is presented. Methods are illustrated and compared with Bayesian ordinary binary regression using data from a study of the effectiveness of a smoking cessation program among pregnant women. Regression coefficient estimates are shown to change noticeably when expert prior knowledge and imperfect sensitivity and specificity are incorporated into the model.

KW - Bayes factor

KW - Bayesian

KW - Logistic regression

KW - Markov chain Monte Carlo

KW - Sensitivity

KW - Specificity

UR - http://www.scopus.com/inward/record.url?scp=1842730232&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=1842730232&partnerID=8YFLogxK

U2 - 10.1002/sim.1656

DO - 10.1002/sim.1656

M3 - Article

C2 - 15057880

AN - SCOPUS:1842730232

VL - 23

SP - 1095

EP - 1109

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 7

ER -