Multiple imputation for early stopping of a complex clinical trial

Research output: Contribution to journalArticle

Abstract

It is desirable to have procedures available for stopping a clinical trial early if there appears to be no treatment effect. Conditional power procedures allow for early stopping in favor of the null hypothesis if the probability of rejecting H0 at the planned end of the trial given the current data and a value of the parameter of interest is below some threshold level. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207- 219) proposed a stochastic curtailment procedure that calculates the conditional power under the alternative hypothesis. Alternatively, predictive power procedures incorporate information from the observed data by averaging the conditional power over the posterior distribution of the parameter. For complex problems in which explicit evaluation of conditional power is not possible, we propose treating the problem of projecting the outcome of a trial given the current data as a missing data problem. We then complete the data using multiple imputation and thus eliminate the need for explicit calculation of conditional power. We apply this method to AIDS Clinical Trials Group (ACTG) protocol 118 and to several simulated clinical trials.

Original languageEnglish (US)
Pages (from-to)229-242
Number of pages14
JournalBiometrics
Volume54
Issue number1
DOIs
StatePublished - Mar 1 1998

Fingerprint

Conditional Power
Early Stopping
Multiple Imputation
Clinical Trials
clinical trials
Statistics
Communication
statistics
Treatment Effects
Clinical Protocols
Posterior distribution
Missing Data
Null hypothesis
Averaging
Acquired Immunodeficiency Syndrome
Eliminate
Calculate
Alternatives
Evaluation
methodology

Keywords

  • Conditional power
  • Predictive power
  • Proportional hazards
  • Stochastic Curtailment
  • Survival analysis

ASJC Scopus subject areas

  • Statistics and Probability
  • Medicine(all)
  • Immunology and Microbiology(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)
  • Applied Mathematics

Cite this

Multiple imputation for early stopping of a complex clinical trial. / Betensky, Rebecca.

In: Biometrics, Vol. 54, No. 1, 01.03.1998, p. 229-242.

Research output: Contribution to journalArticle

@article{8ea49a7108ab4b69b539b59aae61a0dc,
title = "Multiple imputation for early stopping of a complex clinical trial",
abstract = "It is desirable to have procedures available for stopping a clinical trial early if there appears to be no treatment effect. Conditional power procedures allow for early stopping in favor of the null hypothesis if the probability of rejecting H0 at the planned end of the trial given the current data and a value of the parameter of interest is below some threshold level. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207- 219) proposed a stochastic curtailment procedure that calculates the conditional power under the alternative hypothesis. Alternatively, predictive power procedures incorporate information from the observed data by averaging the conditional power over the posterior distribution of the parameter. For complex problems in which explicit evaluation of conditional power is not possible, we propose treating the problem of projecting the outcome of a trial given the current data as a missing data problem. We then complete the data using multiple imputation and thus eliminate the need for explicit calculation of conditional power. We apply this method to AIDS Clinical Trials Group (ACTG) protocol 118 and to several simulated clinical trials.",
keywords = "Conditional power, Predictive power, Proportional hazards, Stochastic Curtailment, Survival analysis",
author = "Rebecca Betensky",
year = "1998",
month = "3",
day = "1",
doi = "10.2307/2534010",
language = "English (US)",
volume = "54",
pages = "229--242",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - Multiple imputation for early stopping of a complex clinical trial

AU - Betensky, Rebecca

PY - 1998/3/1

Y1 - 1998/3/1

N2 - It is desirable to have procedures available for stopping a clinical trial early if there appears to be no treatment effect. Conditional power procedures allow for early stopping in favor of the null hypothesis if the probability of rejecting H0 at the planned end of the trial given the current data and a value of the parameter of interest is below some threshold level. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207- 219) proposed a stochastic curtailment procedure that calculates the conditional power under the alternative hypothesis. Alternatively, predictive power procedures incorporate information from the observed data by averaging the conditional power over the posterior distribution of the parameter. For complex problems in which explicit evaluation of conditional power is not possible, we propose treating the problem of projecting the outcome of a trial given the current data as a missing data problem. We then complete the data using multiple imputation and thus eliminate the need for explicit calculation of conditional power. We apply this method to AIDS Clinical Trials Group (ACTG) protocol 118 and to several simulated clinical trials.

AB - It is desirable to have procedures available for stopping a clinical trial early if there appears to be no treatment effect. Conditional power procedures allow for early stopping in favor of the null hypothesis if the probability of rejecting H0 at the planned end of the trial given the current data and a value of the parameter of interest is below some threshold level. Lan, Simon, and Halperin (1982, Communications in Statistics C1, 207- 219) proposed a stochastic curtailment procedure that calculates the conditional power under the alternative hypothesis. Alternatively, predictive power procedures incorporate information from the observed data by averaging the conditional power over the posterior distribution of the parameter. For complex problems in which explicit evaluation of conditional power is not possible, we propose treating the problem of projecting the outcome of a trial given the current data as a missing data problem. We then complete the data using multiple imputation and thus eliminate the need for explicit calculation of conditional power. We apply this method to AIDS Clinical Trials Group (ACTG) protocol 118 and to several simulated clinical trials.

KW - Conditional power

KW - Predictive power

KW - Proportional hazards

KW - Stochastic Curtailment

KW - Survival analysis

UR - http://www.scopus.com/inward/record.url?scp=0031946425&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031946425&partnerID=8YFLogxK

U2 - 10.2307/2534010

DO - 10.2307/2534010

M3 - Article

C2 - 9544518

AN - SCOPUS:0031946425

VL - 54

SP - 229

EP - 242

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 1

ER -