A Bayesian nonparametric causal inference model for synthesizing randomized clinical trial and real-world evidence

Research output: Contribution to journalArticle

Abstract

With the wide availability of various real-world data (RWD), there is an increasing interest in synthesizing information from both randomized clinical trials and RWD for health-care decision makings. The task of addressing study-specific heterogeneities is one of the most difficult challenges in synthesizing data from disparate sources. Bayesian hierarchical models with nonparametric extension provide a powerful and convenient platform that formalizes the information borrowing strength across the sources. In this paper, we propose a propensity score-based Bayesian nonparametric Dirichlet process mixture model that summarizes subject-level information from randomized and registry studies to draw inference on the causal treatment effect. Simulation studies are conducted to evaluate the model performance under different scenarios. In addition, we demonstrate the proposed method using data from a clinical study on angiotensin converting enzyme inhibitor for treating congestive heart failure.

Original languageEnglish (US)
Pages (from-to)2573-2588
Number of pages16
JournalStatistics in Medicine
Volume38
Issue number14
DOIs
StatePublished - Jun 30 2019

Fingerprint

Nonparametric Inference
Causal Inference
Randomized Clinical Trial
Bayesian Nonparametrics
Propensity Score
Information Storage and Retrieval
Angiotensin-Converting Enzyme Inhibitors
Registries
Decision Making
Randomized Controlled Trials
Heart Failure
Delivery of Health Care
Angiotensin
Congestive Heart Failure
Bayesian Hierarchical Model
Causal Effect
Dirichlet Process
Treatment Effects
Performance Model
Mixture Model

Keywords

  • Bayesian nonparametric hierarchical model
  • causal inference
  • propensity score
  • randomized clinical trial
  • real-world evidence

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

@article{52537f1c1c514f0abcd99bce1617be85,
title = "A Bayesian nonparametric causal inference model for synthesizing randomized clinical trial and real-world evidence",
abstract = "With the wide availability of various real-world data (RWD), there is an increasing interest in synthesizing information from both randomized clinical trials and RWD for health-care decision makings. The task of addressing study-specific heterogeneities is one of the most difficult challenges in synthesizing data from disparate sources. Bayesian hierarchical models with nonparametric extension provide a powerful and convenient platform that formalizes the information borrowing strength across the sources. In this paper, we propose a propensity score-based Bayesian nonparametric Dirichlet process mixture model that summarizes subject-level information from randomized and registry studies to draw inference on the causal treatment effect. Simulation studies are conducted to evaluate the model performance under different scenarios. In addition, we demonstrate the proposed method using data from a clinical study on angiotensin converting enzyme inhibitor for treating congestive heart failure.",
keywords = "Bayesian nonparametric hierarchical model, causal inference, propensity score, randomized clinical trial, real-world evidence",
author = "Chenguang Wang and Gary Rosner",
year = "2019",
month = "6",
day = "30",
doi = "10.1002/sim.8134",
language = "English (US)",
volume = "38",
pages = "2573--2588",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "14",

}

TY - JOUR

T1 - A Bayesian nonparametric causal inference model for synthesizing randomized clinical trial and real-world evidence

AU - Wang, Chenguang

AU - Rosner, Gary

PY - 2019/6/30

Y1 - 2019/6/30

N2 - With the wide availability of various real-world data (RWD), there is an increasing interest in synthesizing information from both randomized clinical trials and RWD for health-care decision makings. The task of addressing study-specific heterogeneities is one of the most difficult challenges in synthesizing data from disparate sources. Bayesian hierarchical models with nonparametric extension provide a powerful and convenient platform that formalizes the information borrowing strength across the sources. In this paper, we propose a propensity score-based Bayesian nonparametric Dirichlet process mixture model that summarizes subject-level information from randomized and registry studies to draw inference on the causal treatment effect. Simulation studies are conducted to evaluate the model performance under different scenarios. In addition, we demonstrate the proposed method using data from a clinical study on angiotensin converting enzyme inhibitor for treating congestive heart failure.

AB - With the wide availability of various real-world data (RWD), there is an increasing interest in synthesizing information from both randomized clinical trials and RWD for health-care decision makings. The task of addressing study-specific heterogeneities is one of the most difficult challenges in synthesizing data from disparate sources. Bayesian hierarchical models with nonparametric extension provide a powerful and convenient platform that formalizes the information borrowing strength across the sources. In this paper, we propose a propensity score-based Bayesian nonparametric Dirichlet process mixture model that summarizes subject-level information from randomized and registry studies to draw inference on the causal treatment effect. Simulation studies are conducted to evaluate the model performance under different scenarios. In addition, we demonstrate the proposed method using data from a clinical study on angiotensin converting enzyme inhibitor for treating congestive heart failure.

KW - Bayesian nonparametric hierarchical model

KW - causal inference

KW - propensity score

KW - randomized clinical trial

KW - real-world evidence

UR - http://www.scopus.com/inward/record.url?scp=85066820297&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066820297&partnerID=8YFLogxK

U2 - 10.1002/sim.8134

DO - 10.1002/sim.8134

M3 - Article

VL - 38

SP - 2573

EP - 2588

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 14

ER -