Prediction intervals for all of M future observations based on linear random effects models

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Seiten (von - bis)283-308
Seitenumfang26
FachzeitschriftStatistica Neerlandica
Jahrgang76
Ausgabenummer3
Frühes Online-Datum19 Dez. 2021
PublikationsstatusVeröffentlicht - 3 Juli 2022

Abstract

In many pharmaceutical and biomedical applications such as assay validation, assessment of historical control data or the detection of anti-drug antibodies, the calculation and interpretation of prediction intervals (PI) is of interest. The present study provides two novel methods for the calculation of prediction intervals based on linear random effects models and REML estimation. Unlike other REML based PI found in literature, both intervals reflect the uncertainty related with the estimation of the prediction variance. The first PI is based on Satterthwaite approximation. For the other PI, a bootstrap calibration approach that we will call quantile-calibration was used. Due to the calibration process this PI can be easily computed for more than one future observation and based on balanced and unbalanced data as well. In order to compare the coverage probabilities of the proposed PI with those of four intervals found in literature, Monte Carlo simulations were run for two relatively complex random effects models and a broad range of parameter settings. The quantile-calibrated PI was implemented in the statistical software R and is available in the predint package

ASJC Scopus Sachgebiete

Zitieren

Prediction intervals for all of M future observations based on linear random effects models. / Menssen, Max; Schaarschmidt, Frank.
in: Statistica Neerlandica, Jahrgang 76, Nr. 3, 03.07.2022, S. 283-308.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Download
@article{1c1c6a51cb9347ad8b3bf05090f8a90f,
title = "Prediction intervals for all of M future observations based on linear random effects models",
abstract = "In many pharmaceutical and biomedical applications such as assay validation, assessment of historical control data or the detection of anti-drug antibodies, the calculation and interpretation of prediction intervals (PI) is of interest. The present study provides two novel methods for the calculation of prediction intervals based on linear random effects models and REML estimation. Unlike other REML based PI found in literature, both intervals reflect the uncertainty related with the estimation of the prediction variance. The first PI is based on Satterthwaite approximation. For the other PI, a bootstrap calibration approach that we will call quantile-calibration was used. Due to the calibration process this PI can be easily computed for more than one future observation and based on balanced and unbalanced data as well. In order to compare the coverage probabilities of the proposed PI with those of four intervals found in literature, Monte Carlo simulations were run for two relatively complex random effects models and a broad range of parameter settings. The quantile-calibrated PI was implemented in the statistical software R and is available in the predint package",
keywords = "Satterthwaite approximation, anti-drug antibody, assay qualification, bootstrap calibration, historical control data",
author = "Max Menssen and Frank Schaarschmidt",
note = "Funding Information: The authors have to thank Clemens Buczilowski for his technical support and Olaf Menssen for fruitful discussions. Furthermore we have to thank the reviewers for reading the manuscript and for their helpful comments. Open Access funding enabled and organized by Projekt DEAL.",
year = "2022",
month = jul,
day = "3",
doi = "https://doi.org/10.1111/stan.12260",
language = "English",
volume = "76",
pages = "283--308",
journal = "Statistica Neerlandica",
issn = "0039-0402",
publisher = "Wiley-Blackwell Publishing Ltd",
number = "3",

}

Download

TY - JOUR

T1 - Prediction intervals for all of M future observations based on linear random effects models

AU - Menssen, Max

AU - Schaarschmidt, Frank

N1 - Funding Information: The authors have to thank Clemens Buczilowski for his technical support and Olaf Menssen for fruitful discussions. Furthermore we have to thank the reviewers for reading the manuscript and for their helpful comments. Open Access funding enabled and organized by Projekt DEAL.

PY - 2022/7/3

Y1 - 2022/7/3

N2 - In many pharmaceutical and biomedical applications such as assay validation, assessment of historical control data or the detection of anti-drug antibodies, the calculation and interpretation of prediction intervals (PI) is of interest. The present study provides two novel methods for the calculation of prediction intervals based on linear random effects models and REML estimation. Unlike other REML based PI found in literature, both intervals reflect the uncertainty related with the estimation of the prediction variance. The first PI is based on Satterthwaite approximation. For the other PI, a bootstrap calibration approach that we will call quantile-calibration was used. Due to the calibration process this PI can be easily computed for more than one future observation and based on balanced and unbalanced data as well. In order to compare the coverage probabilities of the proposed PI with those of four intervals found in literature, Monte Carlo simulations were run for two relatively complex random effects models and a broad range of parameter settings. The quantile-calibrated PI was implemented in the statistical software R and is available in the predint package

AB - In many pharmaceutical and biomedical applications such as assay validation, assessment of historical control data or the detection of anti-drug antibodies, the calculation and interpretation of prediction intervals (PI) is of interest. The present study provides two novel methods for the calculation of prediction intervals based on linear random effects models and REML estimation. Unlike other REML based PI found in literature, both intervals reflect the uncertainty related with the estimation of the prediction variance. The first PI is based on Satterthwaite approximation. For the other PI, a bootstrap calibration approach that we will call quantile-calibration was used. Due to the calibration process this PI can be easily computed for more than one future observation and based on balanced and unbalanced data as well. In order to compare the coverage probabilities of the proposed PI with those of four intervals found in literature, Monte Carlo simulations were run for two relatively complex random effects models and a broad range of parameter settings. The quantile-calibrated PI was implemented in the statistical software R and is available in the predint package

KW - Satterthwaite approximation

KW - anti-drug antibody

KW - assay qualification

KW - bootstrap calibration

KW - historical control data

UR - http://www.scopus.com/inward/record.url?scp=85122785428&partnerID=8YFLogxK

U2 - https://doi.org/10.1111/stan.12260

DO - https://doi.org/10.1111/stan.12260

M3 - Article

VL - 76

SP - 283

EP - 308

JO - Statistica Neerlandica

JF - Statistica Neerlandica

SN - 0039-0402

IS - 3

ER -

Von denselben Autoren