Gold standards are out and Bayes is in: Implementing the cure for imperfect reference tests in diagnostic accuracy studies

Wesley O. Johnson, Geoff Jones, Ian Gardner

Research output: Contribution to journalArticle

Abstract

Bayesian mixture models, often termed latent class models, allow users to estimate the diagnostic accuracy of tests and true prevalence in one or more populations when the positive and/or negative reference standards are imperfect. Moreover, they allow the data analyst to show the superiority of a novel test over an old test, even if this old test is the (imperfect) reference standard. We use published data on Toxoplasmosis in pigs to explore the effects of numbers of tests, numbers of populations, and dependence structure among tests to ensure model (local) identifiability. We discuss and make recommendations about use of priors, sensitivity analysis, model identifiability and study design options, and strongly argue for the use of Bayesian mixture models as a logical and coherent approach for estimating the diagnostic accuracy of two or more tests.

Original languageEnglish (US)
Pages (from-to)113-127
Number of pages15
JournalPreventive Veterinary Medicine
Volume167
DOIs
StatePublished - Jun 1 2019
Externally publishedYes

Fingerprint

Routine Diagnostic Tests
diagnostic techniques
gold
Toxoplasmosis
Population
Swine
testing
reference standards
toxoplasmosis
experimental design
swine

Keywords

  • Bayesian
  • Diagnostic sensitivity and specificity
  • Mixture models
  • Toxoplasmosis

ASJC Scopus subject areas

  • Food Animals
  • Animal Science and Zoology

Cite this

Gold standards are out and Bayes is in : Implementing the cure for imperfect reference tests in diagnostic accuracy studies. / Johnson, Wesley O.; Jones, Geoff; Gardner, Ian.

In: Preventive Veterinary Medicine, Vol. 167, 01.06.2019, p. 113-127.

Research output: Contribution to journalArticle

@article{f21c4b58628142ebac10836d31fd4ef7,
title = "Gold standards are out and Bayes is in: Implementing the cure for imperfect reference tests in diagnostic accuracy studies",
abstract = "Bayesian mixture models, often termed latent class models, allow users to estimate the diagnostic accuracy of tests and true prevalence in one or more populations when the positive and/or negative reference standards are imperfect. Moreover, they allow the data analyst to show the superiority of a novel test over an old test, even if this old test is the (imperfect) reference standard. We use published data on Toxoplasmosis in pigs to explore the effects of numbers of tests, numbers of populations, and dependence structure among tests to ensure model (local) identifiability. We discuss and make recommendations about use of priors, sensitivity analysis, model identifiability and study design options, and strongly argue for the use of Bayesian mixture models as a logical and coherent approach for estimating the diagnostic accuracy of two or more tests.",
keywords = "Bayesian, Diagnostic sensitivity and specificity, Mixture models, Toxoplasmosis",
author = "Johnson, {Wesley O.} and Geoff Jones and Ian Gardner",
year = "2019",
month = "6",
day = "1",
doi = "10.1016/j.prevetmed.2019.01.010",
language = "English (US)",
volume = "167",
pages = "113--127",
journal = "Preventive Veterinary Medicine",
issn = "0167-5877",
publisher = "Elsevier",

}

TY - JOUR

T1 - Gold standards are out and Bayes is in

T2 - Implementing the cure for imperfect reference tests in diagnostic accuracy studies

AU - Johnson, Wesley O.

AU - Jones, Geoff

AU - Gardner, Ian

PY - 2019/6/1

Y1 - 2019/6/1

N2 - Bayesian mixture models, often termed latent class models, allow users to estimate the diagnostic accuracy of tests and true prevalence in one or more populations when the positive and/or negative reference standards are imperfect. Moreover, they allow the data analyst to show the superiority of a novel test over an old test, even if this old test is the (imperfect) reference standard. We use published data on Toxoplasmosis in pigs to explore the effects of numbers of tests, numbers of populations, and dependence structure among tests to ensure model (local) identifiability. We discuss and make recommendations about use of priors, sensitivity analysis, model identifiability and study design options, and strongly argue for the use of Bayesian mixture models as a logical and coherent approach for estimating the diagnostic accuracy of two or more tests.

AB - Bayesian mixture models, often termed latent class models, allow users to estimate the diagnostic accuracy of tests and true prevalence in one or more populations when the positive and/or negative reference standards are imperfect. Moreover, they allow the data analyst to show the superiority of a novel test over an old test, even if this old test is the (imperfect) reference standard. We use published data on Toxoplasmosis in pigs to explore the effects of numbers of tests, numbers of populations, and dependence structure among tests to ensure model (local) identifiability. We discuss and make recommendations about use of priors, sensitivity analysis, model identifiability and study design options, and strongly argue for the use of Bayesian mixture models as a logical and coherent approach for estimating the diagnostic accuracy of two or more tests.

KW - Bayesian

KW - Diagnostic sensitivity and specificity

KW - Mixture models

KW - Toxoplasmosis

UR - http://www.scopus.com/inward/record.url?scp=85064150294&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064150294&partnerID=8YFLogxK

U2 - 10.1016/j.prevetmed.2019.01.010

DO - 10.1016/j.prevetmed.2019.01.010

M3 - Article

C2 - 31027713

AN - SCOPUS:85064150294

VL - 167

SP - 113

EP - 127

JO - Preventive Veterinary Medicine

JF - Preventive Veterinary Medicine

SN - 0167-5877

ER -