Bayesian monitoring of clinical trials with failure-time endpoints

Research output: Contribution to journalArticle

Abstract

This article presents an aid for monitoring clinical trials with failure-time endpoints based on the Bayesian nonparametric analyses of the data. The posterior distribution is a mixture of Dirichlet processes in the presence of censoring if one assumes a Dirichlet process prior for the survival distribution. Using Gibbs sampling, one can generate random samples from the posterior distribution. With samples from the posterior distributions of treatment-specific survival curves, one can evaluate the current evidence in favor of stopping or continuing the trial based on summary statistics of these survival curves. Because the method is nonparametric, it can easily be used, for example, in situations where hazards cross or are suspected to cross and where relevant clinical decisions might be based on estimating when the integral between the curves might be expected to become positive and in favor of the new but toxic therapy. An example based on an actual trial illustrates the method.

Original languageEnglish (US)
Pages (from-to)239-245
Number of pages7
JournalBiometrics
Volume61
Issue number1
DOIs
StatePublished - Mar 2005
Externally publishedYes

Fingerprint

Poisons
Failure Time
endpoints
Posterior distribution
Clinical Trials
clinical trials
Hazards
Statistics
Monitoring
Sampling
Curve
Bayes Theorem
monitoring
Mixture of Dirichlet Processes
Dirichlet Process Prior
Survival Distribution
Bayesian Nonparametrics
Gibbs Sampling
Censoring
sampling

Keywords

  • Clinical trials
  • Dirichlet process
  • Gibbs sampling
  • Survival analysis

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Public Health, Environmental and Occupational Health
  • Agricultural and Biological Sciences (miscellaneous)
  • Applied Mathematics
  • Statistics and Probability

Cite this

Bayesian monitoring of clinical trials with failure-time endpoints. / Rosner, Gary.

In: Biometrics, Vol. 61, No. 1, 03.2005, p. 239-245.

Research output: Contribution to journalArticle

@article{7807be09d50d467e933373fcb089912f,
title = "Bayesian monitoring of clinical trials with failure-time endpoints",
abstract = "This article presents an aid for monitoring clinical trials with failure-time endpoints based on the Bayesian nonparametric analyses of the data. The posterior distribution is a mixture of Dirichlet processes in the presence of censoring if one assumes a Dirichlet process prior for the survival distribution. Using Gibbs sampling, one can generate random samples from the posterior distribution. With samples from the posterior distributions of treatment-specific survival curves, one can evaluate the current evidence in favor of stopping or continuing the trial based on summary statistics of these survival curves. Because the method is nonparametric, it can easily be used, for example, in situations where hazards cross or are suspected to cross and where relevant clinical decisions might be based on estimating when the integral between the curves might be expected to become positive and in favor of the new but toxic therapy. An example based on an actual trial illustrates the method.",
keywords = "Clinical trials, Dirichlet process, Gibbs sampling, Survival analysis",
author = "Gary Rosner",
year = "2005",
month = "3",
doi = "10.1111/j.0006-341X.2005.031037.x",
language = "English (US)",
volume = "61",
pages = "239--245",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - Bayesian monitoring of clinical trials with failure-time endpoints

AU - Rosner, Gary

PY - 2005/3

Y1 - 2005/3

N2 - This article presents an aid for monitoring clinical trials with failure-time endpoints based on the Bayesian nonparametric analyses of the data. The posterior distribution is a mixture of Dirichlet processes in the presence of censoring if one assumes a Dirichlet process prior for the survival distribution. Using Gibbs sampling, one can generate random samples from the posterior distribution. With samples from the posterior distributions of treatment-specific survival curves, one can evaluate the current evidence in favor of stopping or continuing the trial based on summary statistics of these survival curves. Because the method is nonparametric, it can easily be used, for example, in situations where hazards cross or are suspected to cross and where relevant clinical decisions might be based on estimating when the integral between the curves might be expected to become positive and in favor of the new but toxic therapy. An example based on an actual trial illustrates the method.

AB - This article presents an aid for monitoring clinical trials with failure-time endpoints based on the Bayesian nonparametric analyses of the data. The posterior distribution is a mixture of Dirichlet processes in the presence of censoring if one assumes a Dirichlet process prior for the survival distribution. Using Gibbs sampling, one can generate random samples from the posterior distribution. With samples from the posterior distributions of treatment-specific survival curves, one can evaluate the current evidence in favor of stopping or continuing the trial based on summary statistics of these survival curves. Because the method is nonparametric, it can easily be used, for example, in situations where hazards cross or are suspected to cross and where relevant clinical decisions might be based on estimating when the integral between the curves might be expected to become positive and in favor of the new but toxic therapy. An example based on an actual trial illustrates the method.

KW - Clinical trials

KW - Dirichlet process

KW - Gibbs sampling

KW - Survival analysis

UR - http://www.scopus.com/inward/record.url?scp=15044360995&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=15044360995&partnerID=8YFLogxK

U2 - 10.1111/j.0006-341X.2005.031037.x

DO - 10.1111/j.0006-341X.2005.031037.x

M3 - Article

C2 - 15737099

AN - SCOPUS:15044360995

VL - 61

SP - 239

EP - 245

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 1

ER -