A unification of models for meta-analysis of diagnostic accuracy studies without a gold standard

Yulun Liu, Yong Chen, Haitao Chu

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Several statistical methods for meta-analysis of diagnostic accuracy studies have been discussed in the presence of a gold standard. However, in practice, the selected reference test may be imperfect due to measurement error, non-existence, invasive nature, or expensive cost of a gold standard. It has been suggested that treating an imperfect reference test as a gold standard can lead to substantial bias in the estimation of diagnostic test accuracy. Recently, two models have been proposed to account for imperfect reference test, namely, a multivariate generalized linear mixed model (MGLMM) and a hierarchical summary receiver operating characteristic (HSROC) model. Both models are very flexible in accounting for heterogeneity in accuracies of tests across studies as well as the dependence between tests. In this article, we show that these two models, although with different formulations, are closely related and are equivalent in the absence of study-level covariates. Furthermore, we provide the exact relations between the parameters of these two models and assumptions under which two models can be reduced to equivalent submodels. On the other hand, we show that some submodels of the MGLMM do not have corresponding equivalent submodels of the HSROC model, and vice versa. With three real examples, we illustrate the cases when fitting the MGLMM and HSROC models leads to equivalent submodels and hence identical inference, and the cases when the inferences from two models are slightly different. Our results generalize the important relations between the bivariate generalized linear mixed model and HSROC model when the reference test is a gold standard.

Original languageEnglish (US)
Pages (from-to)538-547
Number of pages10
JournalBiometrics
Volume71
Issue number2
DOIs
StatePublished - Jun 1 2015
Externally publishedYes

Fingerprint

Diagnostic Accuracy
Unification
meta-analysis
ROC Curve
Gold
gold
Meta-Analysis
Linear Models
Generalized Linear Mixed Model
Operating Characteristics
Receiver
Imperfect
Statistical Models
Routine Diagnostic Tests
Model
Costs and Cost Analysis
testing
Standards
Diagnostic Tests
Measurement Error

Keywords

  • Diagnostic test
  • Generalized linear mixed model
  • Hierarchical model
  • Imperfect reference test
  • Meta-analysis

ASJC Scopus subject areas

  • Statistics and Probability
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Agricultural and Biological Sciences(all)
  • Applied Mathematics

Cite this

A unification of models for meta-analysis of diagnostic accuracy studies without a gold standard. / Liu, Yulun; Chen, Yong; Chu, Haitao.

In: Biometrics, Vol. 71, No. 2, 01.06.2015, p. 538-547.

Research output: Contribution to journalArticle

@article{f636c3c436cd451094c96b03728b67fa,
title = "A unification of models for meta-analysis of diagnostic accuracy studies without a gold standard",
abstract = "Several statistical methods for meta-analysis of diagnostic accuracy studies have been discussed in the presence of a gold standard. However, in practice, the selected reference test may be imperfect due to measurement error, non-existence, invasive nature, or expensive cost of a gold standard. It has been suggested that treating an imperfect reference test as a gold standard can lead to substantial bias in the estimation of diagnostic test accuracy. Recently, two models have been proposed to account for imperfect reference test, namely, a multivariate generalized linear mixed model (MGLMM) and a hierarchical summary receiver operating characteristic (HSROC) model. Both models are very flexible in accounting for heterogeneity in accuracies of tests across studies as well as the dependence between tests. In this article, we show that these two models, although with different formulations, are closely related and are equivalent in the absence of study-level covariates. Furthermore, we provide the exact relations between the parameters of these two models and assumptions under which two models can be reduced to equivalent submodels. On the other hand, we show that some submodels of the MGLMM do not have corresponding equivalent submodels of the HSROC model, and vice versa. With three real examples, we illustrate the cases when fitting the MGLMM and HSROC models leads to equivalent submodels and hence identical inference, and the cases when the inferences from two models are slightly different. Our results generalize the important relations between the bivariate generalized linear mixed model and HSROC model when the reference test is a gold standard.",
keywords = "Diagnostic test, Generalized linear mixed model, Hierarchical model, Imperfect reference test, Meta-analysis",
author = "Yulun Liu and Yong Chen and Haitao Chu",
year = "2015",
month = "6",
day = "1",
doi = "10.1111/biom.12264",
language = "English (US)",
volume = "71",
pages = "538--547",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - A unification of models for meta-analysis of diagnostic accuracy studies without a gold standard

AU - Liu, Yulun

AU - Chen, Yong

AU - Chu, Haitao

PY - 2015/6/1

Y1 - 2015/6/1

N2 - Several statistical methods for meta-analysis of diagnostic accuracy studies have been discussed in the presence of a gold standard. However, in practice, the selected reference test may be imperfect due to measurement error, non-existence, invasive nature, or expensive cost of a gold standard. It has been suggested that treating an imperfect reference test as a gold standard can lead to substantial bias in the estimation of diagnostic test accuracy. Recently, two models have been proposed to account for imperfect reference test, namely, a multivariate generalized linear mixed model (MGLMM) and a hierarchical summary receiver operating characteristic (HSROC) model. Both models are very flexible in accounting for heterogeneity in accuracies of tests across studies as well as the dependence between tests. In this article, we show that these two models, although with different formulations, are closely related and are equivalent in the absence of study-level covariates. Furthermore, we provide the exact relations between the parameters of these two models and assumptions under which two models can be reduced to equivalent submodels. On the other hand, we show that some submodels of the MGLMM do not have corresponding equivalent submodels of the HSROC model, and vice versa. With three real examples, we illustrate the cases when fitting the MGLMM and HSROC models leads to equivalent submodels and hence identical inference, and the cases when the inferences from two models are slightly different. Our results generalize the important relations between the bivariate generalized linear mixed model and HSROC model when the reference test is a gold standard.

AB - Several statistical methods for meta-analysis of diagnostic accuracy studies have been discussed in the presence of a gold standard. However, in practice, the selected reference test may be imperfect due to measurement error, non-existence, invasive nature, or expensive cost of a gold standard. It has been suggested that treating an imperfect reference test as a gold standard can lead to substantial bias in the estimation of diagnostic test accuracy. Recently, two models have been proposed to account for imperfect reference test, namely, a multivariate generalized linear mixed model (MGLMM) and a hierarchical summary receiver operating characteristic (HSROC) model. Both models are very flexible in accounting for heterogeneity in accuracies of tests across studies as well as the dependence between tests. In this article, we show that these two models, although with different formulations, are closely related and are equivalent in the absence of study-level covariates. Furthermore, we provide the exact relations between the parameters of these two models and assumptions under which two models can be reduced to equivalent submodels. On the other hand, we show that some submodels of the MGLMM do not have corresponding equivalent submodels of the HSROC model, and vice versa. With three real examples, we illustrate the cases when fitting the MGLMM and HSROC models leads to equivalent submodels and hence identical inference, and the cases when the inferences from two models are slightly different. Our results generalize the important relations between the bivariate generalized linear mixed model and HSROC model when the reference test is a gold standard.

KW - Diagnostic test

KW - Generalized linear mixed model

KW - Hierarchical model

KW - Imperfect reference test

KW - Meta-analysis

UR - http://www.scopus.com/inward/record.url?scp=84931348081&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84931348081&partnerID=8YFLogxK

U2 - 10.1111/biom.12264

DO - 10.1111/biom.12264

M3 - Article

C2 - 25358907

AN - SCOPUS:84931348081

VL - 71

SP - 538

EP - 547

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 2

ER -