Simultaneous small-sample comparisons in longitudinal or multi-endpoint trials using multiple marginal models

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Philip Pallmann
  • Christian Ritz
  • Ludwig A. Hothorn

Research Organisations

External Research Organisations

  • Lancaster University
  • University of Copenhagen
View graph of relations

Details

Original languageEnglish
Pages (from-to)1562-1576
Number of pages15
JournalStatistics in medicine
Volume37
Issue number9
Publication statusPublished - 6 Apr 2018

Abstract

Simultaneous inference in longitudinal, repeated-measures, and multi-endpoint designs can be onerous, especially when trying to find a reasonable joint model from which the interesting effects and covariances are estimated. A novel statistical approach known as multiple marginal models greatly simplifies the modelling process: the core idea is to “marginalise” the problem and fit multiple small models to different portions of the data, and then estimate the overall covariance matrix in a subsequent, separate step. Using these estimates guarantees strong control of the family-wise error rate, however only asymptotically. In this paper, we show how to make the approach also applicable to small-sample data problems. Specifically, we discuss the computation of adjusted P values and simultaneous confidence bounds for comparisons of randomised treatment groups as well as for levels of a nonrandomised factor such as multiple endpoints, repeated measures, or a series of points in time or space. We illustrate the practical use of the method with a data example.

Keywords

    correlated data, degrees of freedom, linear mixed-effects model, multiple contrast test

ASJC Scopus subject areas

Cite this

Simultaneous small-sample comparisons in longitudinal or multi-endpoint trials using multiple marginal models. / Pallmann, Philip; Ritz, Christian; Hothorn, Ludwig A.
In: Statistics in medicine, Vol. 37, No. 9, 06.04.2018, p. 1562-1576.

Research output: Contribution to journalArticleResearchpeer review

Pallmann, Philip ; Ritz, Christian ; Hothorn, Ludwig A. / Simultaneous small-sample comparisons in longitudinal or multi-endpoint trials using multiple marginal models. In: Statistics in medicine. 2018 ; Vol. 37, No. 9. pp. 1562-1576.
Download
@article{0df1abc93bea459f90ce3a7e826c94da,
title = "Simultaneous small-sample comparisons in longitudinal or multi-endpoint trials using multiple marginal models",
abstract = "Simultaneous inference in longitudinal, repeated-measures, and multi-endpoint designs can be onerous, especially when trying to find a reasonable joint model from which the interesting effects and covariances are estimated. A novel statistical approach known as multiple marginal models greatly simplifies the modelling process: the core idea is to “marginalise” the problem and fit multiple small models to different portions of the data, and then estimate the overall covariance matrix in a subsequent, separate step. Using these estimates guarantees strong control of the family-wise error rate, however only asymptotically. In this paper, we show how to make the approach also applicable to small-sample data problems. Specifically, we discuss the computation of adjusted P values and simultaneous confidence bounds for comparisons of randomised treatment groups as well as for levels of a nonrandomised factor such as multiple endpoints, repeated measures, or a series of points in time or space. We illustrate the practical use of the method with a data example.",
keywords = "correlated data, degrees of freedom, linear mixed-effects model, multiple contrast test",
author = "Philip Pallmann and Christian Ritz and Hothorn, {Ludwig A.}",
year = "2018",
month = apr,
day = "6",
doi = "10.1002/sim.7610",
language = "English",
volume = "37",
pages = "1562--1576",
journal = "Statistics in medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "9",

}

Download

TY - JOUR

T1 - Simultaneous small-sample comparisons in longitudinal or multi-endpoint trials using multiple marginal models

AU - Pallmann, Philip

AU - Ritz, Christian

AU - Hothorn, Ludwig A.

PY - 2018/4/6

Y1 - 2018/4/6

N2 - Simultaneous inference in longitudinal, repeated-measures, and multi-endpoint designs can be onerous, especially when trying to find a reasonable joint model from which the interesting effects and covariances are estimated. A novel statistical approach known as multiple marginal models greatly simplifies the modelling process: the core idea is to “marginalise” the problem and fit multiple small models to different portions of the data, and then estimate the overall covariance matrix in a subsequent, separate step. Using these estimates guarantees strong control of the family-wise error rate, however only asymptotically. In this paper, we show how to make the approach also applicable to small-sample data problems. Specifically, we discuss the computation of adjusted P values and simultaneous confidence bounds for comparisons of randomised treatment groups as well as for levels of a nonrandomised factor such as multiple endpoints, repeated measures, or a series of points in time or space. We illustrate the practical use of the method with a data example.

AB - Simultaneous inference in longitudinal, repeated-measures, and multi-endpoint designs can be onerous, especially when trying to find a reasonable joint model from which the interesting effects and covariances are estimated. A novel statistical approach known as multiple marginal models greatly simplifies the modelling process: the core idea is to “marginalise” the problem and fit multiple small models to different portions of the data, and then estimate the overall covariance matrix in a subsequent, separate step. Using these estimates guarantees strong control of the family-wise error rate, however only asymptotically. In this paper, we show how to make the approach also applicable to small-sample data problems. Specifically, we discuss the computation of adjusted P values and simultaneous confidence bounds for comparisons of randomised treatment groups as well as for levels of a nonrandomised factor such as multiple endpoints, repeated measures, or a series of points in time or space. We illustrate the practical use of the method with a data example.

KW - correlated data

KW - degrees of freedom

KW - linear mixed-effects model

KW - multiple contrast test

UR - http://www.scopus.com/inward/record.url?scp=85045030872&partnerID=8YFLogxK

U2 - 10.1002/sim.7610

DO - 10.1002/sim.7610

M3 - Article

C2 - 29444546

AN - SCOPUS:85045030872

VL - 37

SP - 1562

EP - 1576

JO - Statistics in medicine

JF - Statistics in medicine

SN - 0277-6715

IS - 9

ER -