Bayesian interim analysis of phase II cancer clinical trials

Research output: Contribution to journalArticle

61 Citations (Scopus)

Abstract

Many popular sequential phase II clinical trial designs optimize some criterion subject to constraints on the error probabilities at null and alternative values of the response rate. Such designs may forfeit optimality if one fails to conduct analyses strictly according to plan. Moreover, a decision, say, to accept the experimental therapy at one interim analysis does not necessarily imply the same degree of evidence as the same decision when made at another analysis. I propose an alternative design that bases decisions on the ability of the data to persuade either a sceptic or an enthusiast. My standard of evidence, called the persuasion probability, is based on the Bayesian posterior probability that the experimental treatment is superior to the standard. The design calls for termination at any interim analysis at which an observed persuasion probability exceeds its critical value. I investigate the standards of evidence implied by some frequentist procedures and calculate frequentist properties of persuasion-probability designs.

Original languageEnglish (US)
Pages (from-to)1791-1802
Number of pages12
JournalStatistics in Medicine
Volume16
Issue number16
DOIs
StatePublished - Aug 30 1997

Fingerprint

Interim Analysis
Phase II Clinical Trials
Bayes Theorem
Bayesian Analysis
Clinical Trials
Cancer
Persuasive Communication
Persuasion
Neoplasms
Investigational Therapies
Alternatives
Posterior Probability
Error Probability
Termination
Therapy
Null
Critical value
Optimality
Exceed
Strictly

ASJC Scopus subject areas

  • Epidemiology

Cite this

Bayesian interim analysis of phase II cancer clinical trials. / Heitjan, Daniel F.

In: Statistics in Medicine, Vol. 16, No. 16, 30.08.1997, p. 1791-1802.

Research output: Contribution to journalArticle

@article{8fe1decdf35e435d934e9be6e747dfa0,
title = "Bayesian interim analysis of phase II cancer clinical trials",
abstract = "Many popular sequential phase II clinical trial designs optimize some criterion subject to constraints on the error probabilities at null and alternative values of the response rate. Such designs may forfeit optimality if one fails to conduct analyses strictly according to plan. Moreover, a decision, say, to accept the experimental therapy at one interim analysis does not necessarily imply the same degree of evidence as the same decision when made at another analysis. I propose an alternative design that bases decisions on the ability of the data to persuade either a sceptic or an enthusiast. My standard of evidence, called the persuasion probability, is based on the Bayesian posterior probability that the experimental treatment is superior to the standard. The design calls for termination at any interim analysis at which an observed persuasion probability exceeds its critical value. I investigate the standards of evidence implied by some frequentist procedures and calculate frequentist properties of persuasion-probability designs.",
author = "Heitjan, {Daniel F.}",
year = "1997",
month = "8",
day = "30",
doi = "10.1002/(SICI)1097-0258(19970830)16:16<1791::AID-SIM609>3.0.CO;2-E",
language = "English (US)",
volume = "16",
pages = "1791--1802",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "16",

}

TY - JOUR

T1 - Bayesian interim analysis of phase II cancer clinical trials

AU - Heitjan, Daniel F.

PY - 1997/8/30

Y1 - 1997/8/30

N2 - Many popular sequential phase II clinical trial designs optimize some criterion subject to constraints on the error probabilities at null and alternative values of the response rate. Such designs may forfeit optimality if one fails to conduct analyses strictly according to plan. Moreover, a decision, say, to accept the experimental therapy at one interim analysis does not necessarily imply the same degree of evidence as the same decision when made at another analysis. I propose an alternative design that bases decisions on the ability of the data to persuade either a sceptic or an enthusiast. My standard of evidence, called the persuasion probability, is based on the Bayesian posterior probability that the experimental treatment is superior to the standard. The design calls for termination at any interim analysis at which an observed persuasion probability exceeds its critical value. I investigate the standards of evidence implied by some frequentist procedures and calculate frequentist properties of persuasion-probability designs.

AB - Many popular sequential phase II clinical trial designs optimize some criterion subject to constraints on the error probabilities at null and alternative values of the response rate. Such designs may forfeit optimality if one fails to conduct analyses strictly according to plan. Moreover, a decision, say, to accept the experimental therapy at one interim analysis does not necessarily imply the same degree of evidence as the same decision when made at another analysis. I propose an alternative design that bases decisions on the ability of the data to persuade either a sceptic or an enthusiast. My standard of evidence, called the persuasion probability, is based on the Bayesian posterior probability that the experimental treatment is superior to the standard. The design calls for termination at any interim analysis at which an observed persuasion probability exceeds its critical value. I investigate the standards of evidence implied by some frequentist procedures and calculate frequentist properties of persuasion-probability designs.

UR - http://www.scopus.com/inward/record.url?scp=0030767683&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030767683&partnerID=8YFLogxK

U2 - 10.1002/(SICI)1097-0258(19970830)16:16<1791::AID-SIM609>3.0.CO;2-E

DO - 10.1002/(SICI)1097-0258(19970830)16:16<1791::AID-SIM609>3.0.CO;2-E

M3 - Article

C2 - 9280033

AN - SCOPUS:0030767683

VL - 16

SP - 1791

EP - 1802

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 16

ER -