Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables

Wei Yin Loh, Haoda Fu, Michael Man, Victoria Champion, Menggang Yu

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

We describe and evaluate a regression tree algorithm for finding subgroups with differential treatments effects in randomized trials with multivariate outcomes. The data may contain missing values in the outcomes and covariates, and the treatment variable is not limited to two levels. Simulation results show that the regression tree models have unbiased variable selection and the estimates of subgroup treatment effects are approximately unbiased. A bootstrap calibration technique is proposed for constructing confidence intervals for the treatment effects. The method is illustrated with data from a longitudinal study comparing two diabetes drugs and a mammography screening trial comparing two treatments and a control.

Original languageEnglish (US)
Pages (from-to)4837-4855
Number of pages19
JournalStatistics in Medicine
Volume35
Issue number26
DOIs
StatePublished - Nov 20 2016

Fingerprint

Treatment Effects
Mammography
Calibration
Longitudinal Studies
Regression Tree
Subgroup
Confidence Intervals
Pharmaceutical Preparations
Randomized Trial
Diabetes
Longitudinal Study
Missing Values
Tree Algorithms
Variable Selection
Bootstrap
Screening
Confidence interval
Covariates
Drugs
Evaluate

Keywords

  • bootstrap
  • precision medicine
  • randomized trial
  • regression tree
  • unbiased

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables. / Loh, Wei Yin; Fu, Haoda; Man, Michael; Champion, Victoria; Yu, Menggang.

In: Statistics in Medicine, Vol. 35, No. 26, 20.11.2016, p. 4837-4855.

Research output: Contribution to journalArticle

Loh, Wei Yin ; Fu, Haoda ; Man, Michael ; Champion, Victoria ; Yu, Menggang. / Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables. In: Statistics in Medicine. 2016 ; Vol. 35, No. 26. pp. 4837-4855.
@article{07a2fa07d0344d6e873cf01cdd882633,
title = "Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables",
abstract = "We describe and evaluate a regression tree algorithm for finding subgroups with differential treatments effects in randomized trials with multivariate outcomes. The data may contain missing values in the outcomes and covariates, and the treatment variable is not limited to two levels. Simulation results show that the regression tree models have unbiased variable selection and the estimates of subgroup treatment effects are approximately unbiased. A bootstrap calibration technique is proposed for constructing confidence intervals for the treatment effects. The method is illustrated with data from a longitudinal study comparing two diabetes drugs and a mammography screening trial comparing two treatments and a control.",
keywords = "bootstrap, precision medicine, randomized trial, regression tree, unbiased",
author = "Loh, {Wei Yin} and Haoda Fu and Michael Man and Victoria Champion and Menggang Yu",
year = "2016",
month = "11",
day = "20",
doi = "10.1002/sim.7020",
language = "English (US)",
volume = "35",
pages = "4837--4855",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "26",

}

TY - JOUR

T1 - Identification of subgroups with differential treatment effects for longitudinal and multiresponse variables

AU - Loh, Wei Yin

AU - Fu, Haoda

AU - Man, Michael

AU - Champion, Victoria

AU - Yu, Menggang

PY - 2016/11/20

Y1 - 2016/11/20

N2 - We describe and evaluate a regression tree algorithm for finding subgroups with differential treatments effects in randomized trials with multivariate outcomes. The data may contain missing values in the outcomes and covariates, and the treatment variable is not limited to two levels. Simulation results show that the regression tree models have unbiased variable selection and the estimates of subgroup treatment effects are approximately unbiased. A bootstrap calibration technique is proposed for constructing confidence intervals for the treatment effects. The method is illustrated with data from a longitudinal study comparing two diabetes drugs and a mammography screening trial comparing two treatments and a control.

AB - We describe and evaluate a regression tree algorithm for finding subgroups with differential treatments effects in randomized trials with multivariate outcomes. The data may contain missing values in the outcomes and covariates, and the treatment variable is not limited to two levels. Simulation results show that the regression tree models have unbiased variable selection and the estimates of subgroup treatment effects are approximately unbiased. A bootstrap calibration technique is proposed for constructing confidence intervals for the treatment effects. The method is illustrated with data from a longitudinal study comparing two diabetes drugs and a mammography screening trial comparing two treatments and a control.

KW - bootstrap

KW - precision medicine

KW - randomized trial

KW - regression tree

KW - unbiased

UR - http://www.scopus.com/inward/record.url?scp=84991503912&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84991503912&partnerID=8YFLogxK

U2 - 10.1002/sim.7020

DO - 10.1002/sim.7020

M3 - Article

VL - 35

SP - 4837

EP - 4855

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 26

ER -