Simultaneous inference for multiple marginal generalized estimating equation models

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Robin Ristl
  • Ludwig Hothorn
  • Christian Ritz
  • Martin Posch

Research Organisations

External Research Organisations

  • Medical University of Vienna
  • University of Copenhagen
View graph of relations

Details

Original languageEnglish
Pages (from-to)1746-1762
Number of pages17
JournalStatistical Methods in Medical Research
Volume29
Issue number6
Early online date17 Sept 2019
Publication statusPublished - 1 Jun 2020

Abstract

Motivated by small-sample studies in ophthalmology and dermatology, we study the problem of simultaneous inference for multiple endpoints in the presence of repeated observations. We propose a framework in which a generalized estimating equation model is fit for each endpoint marginally, taking into account dependencies within the same subject. The asymptotic joint normality of the stacked vector of marginal estimating equations is used to derive Wald-type simultaneous confidence intervals and hypothesis tests for multiple linear contrasts of regression coefficients of the multiple marginal models. The small sample performance of this approach is improved by a bias adjustment to the estimate of the joint covariance matrix of the regression coefficients from multiple models. As a further small sample improvement a multivariate t-distribution with appropriate degrees of freedom is specified as reference distribution. In addition, a generalized score test based on the stacked estimating equations is derived. Simulation results show strong control of the family-wise type I error rate for these methods even with small sample sizes and increased power compared to a Bonferroni-Holm multiplicity adjustment. Thus, the proposed methods are suitable to efficiently use the information from repeated observations of multiple endpoints in small-sample studies.

Keywords

    dependent observations, Generalized estimating equations, multiple endpoints, multiple testing, small samples

ASJC Scopus subject areas

Cite this

Simultaneous inference for multiple marginal generalized estimating equation models. / Ristl, Robin; Hothorn, Ludwig; Ritz, Christian et al.
In: Statistical Methods in Medical Research, Vol. 29, No. 6, 01.06.2020, p. 1746-1762.

Research output: Contribution to journalArticleResearchpeer review

Ristl R, Hothorn L, Ritz C, Posch M. Simultaneous inference for multiple marginal generalized estimating equation models. Statistical Methods in Medical Research. 2020 Jun 1;29(6):1746-1762. Epub 2019 Sept 17. doi: 10.1177/0962280219873005
Ristl, Robin ; Hothorn, Ludwig ; Ritz, Christian et al. / Simultaneous inference for multiple marginal generalized estimating equation models. In: Statistical Methods in Medical Research. 2020 ; Vol. 29, No. 6. pp. 1746-1762.
Download
@article{d084acf5b32a4d20aac2f8e328382333,
title = "Simultaneous inference for multiple marginal generalized estimating equation models",
abstract = "Motivated by small-sample studies in ophthalmology and dermatology, we study the problem of simultaneous inference for multiple endpoints in the presence of repeated observations. We propose a framework in which a generalized estimating equation model is fit for each endpoint marginally, taking into account dependencies within the same subject. The asymptotic joint normality of the stacked vector of marginal estimating equations is used to derive Wald-type simultaneous confidence intervals and hypothesis tests for multiple linear contrasts of regression coefficients of the multiple marginal models. The small sample performance of this approach is improved by a bias adjustment to the estimate of the joint covariance matrix of the regression coefficients from multiple models. As a further small sample improvement a multivariate t-distribution with appropriate degrees of freedom is specified as reference distribution. In addition, a generalized score test based on the stacked estimating equations is derived. Simulation results show strong control of the family-wise type I error rate for these methods even with small sample sizes and increased power compared to a Bonferroni-Holm multiplicity adjustment. Thus, the proposed methods are suitable to efficiently use the information from repeated observations of multiple endpoints in small-sample studies.",
keywords = "dependent observations, Generalized estimating equations, multiple endpoints, multiple testing, small samples",
author = "Robin Ristl and Ludwig Hothorn and Christian Ritz and Martin Posch",
year = "2020",
month = jun,
day = "1",
doi = "10.1177/0962280219873005",
language = "English",
volume = "29",
pages = "1746--1762",
journal = "Statistical Methods in Medical Research",
issn = "0962-2802",
publisher = "SAGE Publications Ltd",
number = "6",

}

Download

TY - JOUR

T1 - Simultaneous inference for multiple marginal generalized estimating equation models

AU - Ristl, Robin

AU - Hothorn, Ludwig

AU - Ritz, Christian

AU - Posch, Martin

PY - 2020/6/1

Y1 - 2020/6/1

N2 - Motivated by small-sample studies in ophthalmology and dermatology, we study the problem of simultaneous inference for multiple endpoints in the presence of repeated observations. We propose a framework in which a generalized estimating equation model is fit for each endpoint marginally, taking into account dependencies within the same subject. The asymptotic joint normality of the stacked vector of marginal estimating equations is used to derive Wald-type simultaneous confidence intervals and hypothesis tests for multiple linear contrasts of regression coefficients of the multiple marginal models. The small sample performance of this approach is improved by a bias adjustment to the estimate of the joint covariance matrix of the regression coefficients from multiple models. As a further small sample improvement a multivariate t-distribution with appropriate degrees of freedom is specified as reference distribution. In addition, a generalized score test based on the stacked estimating equations is derived. Simulation results show strong control of the family-wise type I error rate for these methods even with small sample sizes and increased power compared to a Bonferroni-Holm multiplicity adjustment. Thus, the proposed methods are suitable to efficiently use the information from repeated observations of multiple endpoints in small-sample studies.

AB - Motivated by small-sample studies in ophthalmology and dermatology, we study the problem of simultaneous inference for multiple endpoints in the presence of repeated observations. We propose a framework in which a generalized estimating equation model is fit for each endpoint marginally, taking into account dependencies within the same subject. The asymptotic joint normality of the stacked vector of marginal estimating equations is used to derive Wald-type simultaneous confidence intervals and hypothesis tests for multiple linear contrasts of regression coefficients of the multiple marginal models. The small sample performance of this approach is improved by a bias adjustment to the estimate of the joint covariance matrix of the regression coefficients from multiple models. As a further small sample improvement a multivariate t-distribution with appropriate degrees of freedom is specified as reference distribution. In addition, a generalized score test based on the stacked estimating equations is derived. Simulation results show strong control of the family-wise type I error rate for these methods even with small sample sizes and increased power compared to a Bonferroni-Holm multiplicity adjustment. Thus, the proposed methods are suitable to efficiently use the information from repeated observations of multiple endpoints in small-sample studies.

KW - dependent observations

KW - Generalized estimating equations

KW - multiple endpoints

KW - multiple testing

KW - small samples

UR - http://www.scopus.com/inward/record.url?scp=85073982153&partnerID=8YFLogxK

U2 - 10.1177/0962280219873005

DO - 10.1177/0962280219873005

M3 - Article

C2 - 31526178

AN - SCOPUS:85073982153

VL - 29

SP - 1746

EP - 1762

JO - Statistical Methods in Medical Research

JF - Statistical Methods in Medical Research

SN - 0962-2802

IS - 6

ER -