A nested Dirichlet process analysis of cluster randomized trial data with application in geriatric care assessment

Man Wai Ho, Wanzhu Tu, Pulak Ghosh, Ram C. Tiwari

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In cluster randomized trials, patients seen by the same physician are randomized to the same treatment arm as a group. Besides the natural clustering of patients due to cluster/group randomization, interactions between an individual patient and the attending physician within the group could just as well influence patient care outcomes. Despite the intuitive relevance of these interactions to treatment assessment, few studies have thus far examined their influences. Whether and to what extent these interactions affect assessment of the treatment effect remains unexplored. In fact, few statistical models provide ready accommodation for such interactions. In this research, we propose a general modeling framework based on the nested Dirichlet process (nDP) for assessing treatment effect in cluster randomized trials. The proposed methodology explicitly accounts for physician-patient interactions by assuming that the interactions follow unspecified group-specific distributions from an nDP. In addition to accounting for physician-patient interactions, the model has greatly enhanced the flexibility of traditional mixed effect models by allowing for nonnormally distributed random effects, thus, alleviating concerns about mixed effect misspecification and sidestepping verification of distributional assumptions on random effects. At the same time, the model retains the mixed models' ability to make inferences on fixed effects. The proposed method is easily extendable to more complicated hierarchical clustering structures. We introduce the method in the context of a real cluster randomized trial. A comprehensive simulation study was conducted to assess the operating characteristics of the proposed nDP model.

Original languageEnglish
Pages (from-to)48-68
Number of pages21
JournalJournal of the American Statistical Association
Volume108
Issue number501
DOIs
StatePublished - 2013

Fingerprint

Randomized Trial
Dirichlet Process
Interaction
Treatment Effects
Random Effects
Mixed Effects
Mixed Effects Model
Fixed Effects
Misspecification
Operating Characteristics
Hierarchical Clustering
Mixed Model
Randomisation
Randomized trial
Process analysis
Dirichlet process
Process Model
Statistical Model
Intuitive
Flexibility

Keywords

  • Clustered data
  • Random effects distribution
  • Treatment effect

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

A nested Dirichlet process analysis of cluster randomized trial data with application in geriatric care assessment. / Ho, Man Wai; Tu, Wanzhu; Ghosh, Pulak; Tiwari, Ram C.

In: Journal of the American Statistical Association, Vol. 108, No. 501, 2013, p. 48-68.

Research output: Contribution to journalArticle

@article{7768bc2ae8b3498aa96594530c0ec8b4,
title = "A nested Dirichlet process analysis of cluster randomized trial data with application in geriatric care assessment",
abstract = "In cluster randomized trials, patients seen by the same physician are randomized to the same treatment arm as a group. Besides the natural clustering of patients due to cluster/group randomization, interactions between an individual patient and the attending physician within the group could just as well influence patient care outcomes. Despite the intuitive relevance of these interactions to treatment assessment, few studies have thus far examined their influences. Whether and to what extent these interactions affect assessment of the treatment effect remains unexplored. In fact, few statistical models provide ready accommodation for such interactions. In this research, we propose a general modeling framework based on the nested Dirichlet process (nDP) for assessing treatment effect in cluster randomized trials. The proposed methodology explicitly accounts for physician-patient interactions by assuming that the interactions follow unspecified group-specific distributions from an nDP. In addition to accounting for physician-patient interactions, the model has greatly enhanced the flexibility of traditional mixed effect models by allowing for nonnormally distributed random effects, thus, alleviating concerns about mixed effect misspecification and sidestepping verification of distributional assumptions on random effects. At the same time, the model retains the mixed models' ability to make inferences on fixed effects. The proposed method is easily extendable to more complicated hierarchical clustering structures. We introduce the method in the context of a real cluster randomized trial. A comprehensive simulation study was conducted to assess the operating characteristics of the proposed nDP model.",
keywords = "Clustered data, Random effects distribution, Treatment effect",
author = "Ho, {Man Wai} and Wanzhu Tu and Pulak Ghosh and Tiwari, {Ram C.}",
year = "2013",
doi = "10.1080/01621459.2012.734164",
language = "English",
volume = "108",
pages = "48--68",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "501",

}

TY - JOUR

T1 - A nested Dirichlet process analysis of cluster randomized trial data with application in geriatric care assessment

AU - Ho, Man Wai

AU - Tu, Wanzhu

AU - Ghosh, Pulak

AU - Tiwari, Ram C.

PY - 2013

Y1 - 2013

N2 - In cluster randomized trials, patients seen by the same physician are randomized to the same treatment arm as a group. Besides the natural clustering of patients due to cluster/group randomization, interactions between an individual patient and the attending physician within the group could just as well influence patient care outcomes. Despite the intuitive relevance of these interactions to treatment assessment, few studies have thus far examined their influences. Whether and to what extent these interactions affect assessment of the treatment effect remains unexplored. In fact, few statistical models provide ready accommodation for such interactions. In this research, we propose a general modeling framework based on the nested Dirichlet process (nDP) for assessing treatment effect in cluster randomized trials. The proposed methodology explicitly accounts for physician-patient interactions by assuming that the interactions follow unspecified group-specific distributions from an nDP. In addition to accounting for physician-patient interactions, the model has greatly enhanced the flexibility of traditional mixed effect models by allowing for nonnormally distributed random effects, thus, alleviating concerns about mixed effect misspecification and sidestepping verification of distributional assumptions on random effects. At the same time, the model retains the mixed models' ability to make inferences on fixed effects. The proposed method is easily extendable to more complicated hierarchical clustering structures. We introduce the method in the context of a real cluster randomized trial. A comprehensive simulation study was conducted to assess the operating characteristics of the proposed nDP model.

AB - In cluster randomized trials, patients seen by the same physician are randomized to the same treatment arm as a group. Besides the natural clustering of patients due to cluster/group randomization, interactions between an individual patient and the attending physician within the group could just as well influence patient care outcomes. Despite the intuitive relevance of these interactions to treatment assessment, few studies have thus far examined their influences. Whether and to what extent these interactions affect assessment of the treatment effect remains unexplored. In fact, few statistical models provide ready accommodation for such interactions. In this research, we propose a general modeling framework based on the nested Dirichlet process (nDP) for assessing treatment effect in cluster randomized trials. The proposed methodology explicitly accounts for physician-patient interactions by assuming that the interactions follow unspecified group-specific distributions from an nDP. In addition to accounting for physician-patient interactions, the model has greatly enhanced the flexibility of traditional mixed effect models by allowing for nonnormally distributed random effects, thus, alleviating concerns about mixed effect misspecification and sidestepping verification of distributional assumptions on random effects. At the same time, the model retains the mixed models' ability to make inferences on fixed effects. The proposed method is easily extendable to more complicated hierarchical clustering structures. We introduce the method in the context of a real cluster randomized trial. A comprehensive simulation study was conducted to assess the operating characteristics of the proposed nDP model.

KW - Clustered data

KW - Random effects distribution

KW - Treatment effect

UR - http://www.scopus.com/inward/record.url?scp=84878258209&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84878258209&partnerID=8YFLogxK

U2 - 10.1080/01621459.2012.734164

DO - 10.1080/01621459.2012.734164

M3 - Article

AN - SCOPUS:84878258209

VL - 108

SP - 48

EP - 68

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 501

ER -