Meta-analysis of a continuous outcome combining individual patient data and aggregate data: A method based on simulated individual patient data

Yusuke Yamaguchi, Wataru Sakamoto, Masashi Goto, Jan A. Staessen, Jiguang Wang, Francois Gueyffier, Richard D. Riley

Research output: Contribution to journalArticle

Abstract

When some trials provide individual patient data (IPD) and the others provide only aggregate data (AD), meta-analysis methods for combining IPD and AD are required. We propose a method that reconstructs the missing IPD for AD trials by a Bayesian sampling procedure and then applies an IPD meta-analysis model to the mixture of simulated IPD and collected IPD. The method is applicable when a treatment effect can be assumed fixed across trials. We focus on situations of a single continuous outcome and covariate and aim to estimate treatment-covariate interactions separated into within-trial and across-trial effect. An illustration with hypertension data which has similar mean covariates across trials indicates that the method substantially reduces mean square error of the pooled within-trial interaction estimate in comparison with existing approaches. A simulation study supposing there exists one IPD trial and nine AD trials suggests that the method has suitable type I error rate and approximately zero bias as long as the available IPD contains at least 10% of total patients, where the average gain in mean square error is up to about 40%. However, the method is currently restricted by the fixed effect assumption, and extension to random effects to allow heterogeneity is required.

Original languageEnglish
Pages (from-to)322-351
Number of pages30
JournalResearch Synthesis Methods
Volume5
Issue number4
DOIs
Publication statusPublished - Dec 1 2014

Fingerprint

aggregate data
model analysis
hypertension
interaction
simulation

Keywords

  • Individual patient data
  • Meta-analysis
  • Statistical simulation
  • Treatment-covariate interaction

ASJC Scopus subject areas

  • Education

Cite this

Meta-analysis of a continuous outcome combining individual patient data and aggregate data : A method based on simulated individual patient data. / Yamaguchi, Yusuke; Sakamoto, Wataru; Goto, Masashi; Staessen, Jan A.; Wang, Jiguang; Gueyffier, Francois; Riley, Richard D.

In: Research Synthesis Methods, Vol. 5, No. 4, 01.12.2014, p. 322-351.

Research output: Contribution to journalArticle

Yamaguchi, Yusuke ; Sakamoto, Wataru ; Goto, Masashi ; Staessen, Jan A. ; Wang, Jiguang ; Gueyffier, Francois ; Riley, Richard D. / Meta-analysis of a continuous outcome combining individual patient data and aggregate data : A method based on simulated individual patient data. In: Research Synthesis Methods. 2014 ; Vol. 5, No. 4. pp. 322-351.
@article{ca7c3a7a358d407986a9c714eb4bc823,
title = "Meta-analysis of a continuous outcome combining individual patient data and aggregate data: A method based on simulated individual patient data",
abstract = "When some trials provide individual patient data (IPD) and the others provide only aggregate data (AD), meta-analysis methods for combining IPD and AD are required. We propose a method that reconstructs the missing IPD for AD trials by a Bayesian sampling procedure and then applies an IPD meta-analysis model to the mixture of simulated IPD and collected IPD. The method is applicable when a treatment effect can be assumed fixed across trials. We focus on situations of a single continuous outcome and covariate and aim to estimate treatment-covariate interactions separated into within-trial and across-trial effect. An illustration with hypertension data which has similar mean covariates across trials indicates that the method substantially reduces mean square error of the pooled within-trial interaction estimate in comparison with existing approaches. A simulation study supposing there exists one IPD trial and nine AD trials suggests that the method has suitable type I error rate and approximately zero bias as long as the available IPD contains at least 10{\%} of total patients, where the average gain in mean square error is up to about 40{\%}. However, the method is currently restricted by the fixed effect assumption, and extension to random effects to allow heterogeneity is required.",
keywords = "Individual patient data, Meta-analysis, Statistical simulation, Treatment-covariate interaction",
author = "Yusuke Yamaguchi and Wataru Sakamoto and Masashi Goto and Staessen, {Jan A.} and Jiguang Wang and Francois Gueyffier and Riley, {Richard D.}",
year = "2014",
month = "12",
day = "1",
doi = "10.1002/jrsm.1119",
language = "English",
volume = "5",
pages = "322--351",
journal = "Research Synthesis Methods",
issn = "1759-2879",
publisher = "John Wiley and Sons Ltd",
number = "4",

}

TY - JOUR

T1 - Meta-analysis of a continuous outcome combining individual patient data and aggregate data

T2 - A method based on simulated individual patient data

AU - Yamaguchi, Yusuke

AU - Sakamoto, Wataru

AU - Goto, Masashi

AU - Staessen, Jan A.

AU - Wang, Jiguang

AU - Gueyffier, Francois

AU - Riley, Richard D.

PY - 2014/12/1

Y1 - 2014/12/1

N2 - When some trials provide individual patient data (IPD) and the others provide only aggregate data (AD), meta-analysis methods for combining IPD and AD are required. We propose a method that reconstructs the missing IPD for AD trials by a Bayesian sampling procedure and then applies an IPD meta-analysis model to the mixture of simulated IPD and collected IPD. The method is applicable when a treatment effect can be assumed fixed across trials. We focus on situations of a single continuous outcome and covariate and aim to estimate treatment-covariate interactions separated into within-trial and across-trial effect. An illustration with hypertension data which has similar mean covariates across trials indicates that the method substantially reduces mean square error of the pooled within-trial interaction estimate in comparison with existing approaches. A simulation study supposing there exists one IPD trial and nine AD trials suggests that the method has suitable type I error rate and approximately zero bias as long as the available IPD contains at least 10% of total patients, where the average gain in mean square error is up to about 40%. However, the method is currently restricted by the fixed effect assumption, and extension to random effects to allow heterogeneity is required.

AB - When some trials provide individual patient data (IPD) and the others provide only aggregate data (AD), meta-analysis methods for combining IPD and AD are required. We propose a method that reconstructs the missing IPD for AD trials by a Bayesian sampling procedure and then applies an IPD meta-analysis model to the mixture of simulated IPD and collected IPD. The method is applicable when a treatment effect can be assumed fixed across trials. We focus on situations of a single continuous outcome and covariate and aim to estimate treatment-covariate interactions separated into within-trial and across-trial effect. An illustration with hypertension data which has similar mean covariates across trials indicates that the method substantially reduces mean square error of the pooled within-trial interaction estimate in comparison with existing approaches. A simulation study supposing there exists one IPD trial and nine AD trials suggests that the method has suitable type I error rate and approximately zero bias as long as the available IPD contains at least 10% of total patients, where the average gain in mean square error is up to about 40%. However, the method is currently restricted by the fixed effect assumption, and extension to random effects to allow heterogeneity is required.

KW - Individual patient data

KW - Meta-analysis

KW - Statistical simulation

KW - Treatment-covariate interaction

UR - http://www.scopus.com/inward/record.url?scp=84959329252&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959329252&partnerID=8YFLogxK

U2 - 10.1002/jrsm.1119

DO - 10.1002/jrsm.1119

M3 - Article

C2 - 26052956

AN - SCOPUS:84959329252

VL - 5

SP - 322

EP - 351

JO - Research Synthesis Methods

JF - Research Synthesis Methods

SN - 1759-2879

IS - 4

ER -