Robust Bayesian approaches for clinical trial monitoring

Bradley P. Carlin, Daniel J. Sargent

Research output: Contribution to journalArticle

21 Citations (Scopus)

Abstract

The interim monitoring and final analysis of data arising from a clinical trial require an inferential method capable of convincing a broad group of potential consumers: doctors; patients; politicians; members of the media, and so on. While Bayesian methods offer a powerful and flexible analytic framework in this setting, this need to convince a diverse community necessitates a practical approach for studying and communicating the robustness of conclusions to the prior specification. In this paper we attempt to characterize the class of priors leading to a given decision (such as stopping the trial and rejecting the null hypothesis) conditional on the observed data. We evaluate the practicality and effectiveness of this procedure over a range of smoothness conditions on the prior class. First, we consider a non-parametric class of priors restricted only in that its elements must have certain prespecified quantiles. We then obtain more precise results by further restricting the prior class, first to a non-parametric class whose members are quasi-unimodal, then to a semi-parametric normal mixture class, and finally to the fully parametric normal family. We illustrate all of our comparisons with a dataset from an AIDS clinical trial that compared the effectiveness of the drug pyrimethamine and a placebo in preventing toxoplasmic encephalitis.

Original languageEnglish (US)
Pages (from-to)1093-1106
Number of pages14
JournalStatistics in Medicine
Volume15
Issue number11
DOIs
StatePublished - 1996
Externally publishedYes

Fingerprint

Bayes Theorem
Bayesian Approach
Clinical Trials
Monitoring
Pyrimethamine
Encephalitis
Acquired Immunodeficiency Syndrome
Placebos
Pharmaceutical Preparations
Normal Family
Normal Mixture
Bayesian Methods
Quantile
Null hypothesis
Class
Smoothness
Drugs
Specification
Robustness
Evaluate

ASJC Scopus subject areas

  • Epidemiology

Cite this

Robust Bayesian approaches for clinical trial monitoring. / Carlin, Bradley P.; Sargent, Daniel J.

In: Statistics in Medicine, Vol. 15, No. 11, 1996, p. 1093-1106.

Research output: Contribution to journalArticle

Carlin, Bradley P. ; Sargent, Daniel J. / Robust Bayesian approaches for clinical trial monitoring. In: Statistics in Medicine. 1996 ; Vol. 15, No. 11. pp. 1093-1106.
@article{dbe505bf52074e46bbc482fa2c8e6231,
title = "Robust Bayesian approaches for clinical trial monitoring",
abstract = "The interim monitoring and final analysis of data arising from a clinical trial require an inferential method capable of convincing a broad group of potential consumers: doctors; patients; politicians; members of the media, and so on. While Bayesian methods offer a powerful and flexible analytic framework in this setting, this need to convince a diverse community necessitates a practical approach for studying and communicating the robustness of conclusions to the prior specification. In this paper we attempt to characterize the class of priors leading to a given decision (such as stopping the trial and rejecting the null hypothesis) conditional on the observed data. We evaluate the practicality and effectiveness of this procedure over a range of smoothness conditions on the prior class. First, we consider a non-parametric class of priors restricted only in that its elements must have certain prespecified quantiles. We then obtain more precise results by further restricting the prior class, first to a non-parametric class whose members are quasi-unimodal, then to a semi-parametric normal mixture class, and finally to the fully parametric normal family. We illustrate all of our comparisons with a dataset from an AIDS clinical trial that compared the effectiveness of the drug pyrimethamine and a placebo in preventing toxoplasmic encephalitis.",
author = "Carlin, {Bradley P.} and Sargent, {Daniel J.}",
year = "1996",
doi = "10.1002/(SICI)1097-0258(19960615)15:11<1093::AID-SIM231>3.0.CO;2-0",
language = "English (US)",
volume = "15",
pages = "1093--1106",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "11",

}

TY - JOUR

T1 - Robust Bayesian approaches for clinical trial monitoring

AU - Carlin, Bradley P.

AU - Sargent, Daniel J.

PY - 1996

Y1 - 1996

N2 - The interim monitoring and final analysis of data arising from a clinical trial require an inferential method capable of convincing a broad group of potential consumers: doctors; patients; politicians; members of the media, and so on. While Bayesian methods offer a powerful and flexible analytic framework in this setting, this need to convince a diverse community necessitates a practical approach for studying and communicating the robustness of conclusions to the prior specification. In this paper we attempt to characterize the class of priors leading to a given decision (such as stopping the trial and rejecting the null hypothesis) conditional on the observed data. We evaluate the practicality and effectiveness of this procedure over a range of smoothness conditions on the prior class. First, we consider a non-parametric class of priors restricted only in that its elements must have certain prespecified quantiles. We then obtain more precise results by further restricting the prior class, first to a non-parametric class whose members are quasi-unimodal, then to a semi-parametric normal mixture class, and finally to the fully parametric normal family. We illustrate all of our comparisons with a dataset from an AIDS clinical trial that compared the effectiveness of the drug pyrimethamine and a placebo in preventing toxoplasmic encephalitis.

AB - The interim monitoring and final analysis of data arising from a clinical trial require an inferential method capable of convincing a broad group of potential consumers: doctors; patients; politicians; members of the media, and so on. While Bayesian methods offer a powerful and flexible analytic framework in this setting, this need to convince a diverse community necessitates a practical approach for studying and communicating the robustness of conclusions to the prior specification. In this paper we attempt to characterize the class of priors leading to a given decision (such as stopping the trial and rejecting the null hypothesis) conditional on the observed data. We evaluate the practicality and effectiveness of this procedure over a range of smoothness conditions on the prior class. First, we consider a non-parametric class of priors restricted only in that its elements must have certain prespecified quantiles. We then obtain more precise results by further restricting the prior class, first to a non-parametric class whose members are quasi-unimodal, then to a semi-parametric normal mixture class, and finally to the fully parametric normal family. We illustrate all of our comparisons with a dataset from an AIDS clinical trial that compared the effectiveness of the drug pyrimethamine and a placebo in preventing toxoplasmic encephalitis.

UR - http://www.scopus.com/inward/record.url?scp=0029898589&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029898589&partnerID=8YFLogxK

U2 - 10.1002/(SICI)1097-0258(19960615)15:11<1093::AID-SIM231>3.0.CO;2-0

DO - 10.1002/(SICI)1097-0258(19960615)15:11<1093::AID-SIM231>3.0.CO;2-0

M3 - Article

C2 - 8804141

AN - SCOPUS:0029898589

VL - 15

SP - 1093

EP - 1106

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 11

ER -