Predicting Treatment Effect from Surrogate Endpoints and Historical Trials: An Extrapolation Involving Probabilities of a Binary Outcome or Survival to a Specific Time

Stuart G. Baker, Daniel J. Sargent, Marc Buyse, Tomasz Burzykowski

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Using multiple historical trials with surrogate and true endpoints, we consider various models to predict the effect of treatment on a true endpoint in a target trial in which only a surrogate endpoint is observed. This predicted result is computed using (1) a prediction model (mixture, linear, or principal stratification) estimated from historical trials and the surrogate endpoint of the target trial and (2) a random extrapolation error estimated from successively leaving out each trial among the historical trials. The method applies to either binary outcomes or survival to a particular time that is computed from censored survival data. We compute a 95% confidence interval for the predicted result and validate its coverage using simulation. To summarize the additional uncertainty from using a predicted instead of true result for the estimated treatment effect, we compute its multiplier of standard error. Software is available for download.

Original languageEnglish (US)
Pages (from-to)248-257
Number of pages10
JournalBiometrics
Volume68
Issue number1
DOIs
StatePublished - Mar 2012

Fingerprint

Surrogate Endpoint
Binary Outcomes
Treatment Effects
endpoints
Extrapolation
Biomarkers
multipliers
Uncertainty
Linear Models
Software
Principal Stratification
Confidence Intervals
Censored Survival Data
confidence interval
Target
uncertainty
Standard error
Prediction Model
Multiplier
prediction

Keywords

  • Principal stratification
  • Randomized trials
  • Reproducibility

ASJC Scopus subject areas

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

Cite this

Predicting Treatment Effect from Surrogate Endpoints and Historical Trials : An Extrapolation Involving Probabilities of a Binary Outcome or Survival to a Specific Time. / Baker, Stuart G.; Sargent, Daniel J.; Buyse, Marc; Burzykowski, Tomasz.

In: Biometrics, Vol. 68, No. 1, 03.2012, p. 248-257.

Research output: Contribution to journalArticle

Baker, Stuart G. ; Sargent, Daniel J. ; Buyse, Marc ; Burzykowski, Tomasz. / Predicting Treatment Effect from Surrogate Endpoints and Historical Trials : An Extrapolation Involving Probabilities of a Binary Outcome or Survival to a Specific Time. In: Biometrics. 2012 ; Vol. 68, No. 1. pp. 248-257.
@article{f0120c912ceb41299cecf1a24f523e4a,
title = "Predicting Treatment Effect from Surrogate Endpoints and Historical Trials: An Extrapolation Involving Probabilities of a Binary Outcome or Survival to a Specific Time",
abstract = "Using multiple historical trials with surrogate and true endpoints, we consider various models to predict the effect of treatment on a true endpoint in a target trial in which only a surrogate endpoint is observed. This predicted result is computed using (1) a prediction model (mixture, linear, or principal stratification) estimated from historical trials and the surrogate endpoint of the target trial and (2) a random extrapolation error estimated from successively leaving out each trial among the historical trials. The method applies to either binary outcomes or survival to a particular time that is computed from censored survival data. We compute a 95{\%} confidence interval for the predicted result and validate its coverage using simulation. To summarize the additional uncertainty from using a predicted instead of true result for the estimated treatment effect, we compute its multiplier of standard error. Software is available for download.",
keywords = "Principal stratification, Randomized trials, Reproducibility",
author = "Baker, {Stuart G.} and Sargent, {Daniel J.} and Marc Buyse and Tomasz Burzykowski",
year = "2012",
month = "3",
doi = "10.1111/j.1541-0420.2011.01646.x",
language = "English (US)",
volume = "68",
pages = "248--257",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - Predicting Treatment Effect from Surrogate Endpoints and Historical Trials

T2 - An Extrapolation Involving Probabilities of a Binary Outcome or Survival to a Specific Time

AU - Baker, Stuart G.

AU - Sargent, Daniel J.

AU - Buyse, Marc

AU - Burzykowski, Tomasz

PY - 2012/3

Y1 - 2012/3

N2 - Using multiple historical trials with surrogate and true endpoints, we consider various models to predict the effect of treatment on a true endpoint in a target trial in which only a surrogate endpoint is observed. This predicted result is computed using (1) a prediction model (mixture, linear, or principal stratification) estimated from historical trials and the surrogate endpoint of the target trial and (2) a random extrapolation error estimated from successively leaving out each trial among the historical trials. The method applies to either binary outcomes or survival to a particular time that is computed from censored survival data. We compute a 95% confidence interval for the predicted result and validate its coverage using simulation. To summarize the additional uncertainty from using a predicted instead of true result for the estimated treatment effect, we compute its multiplier of standard error. Software is available for download.

AB - Using multiple historical trials with surrogate and true endpoints, we consider various models to predict the effect of treatment on a true endpoint in a target trial in which only a surrogate endpoint is observed. This predicted result is computed using (1) a prediction model (mixture, linear, or principal stratification) estimated from historical trials and the surrogate endpoint of the target trial and (2) a random extrapolation error estimated from successively leaving out each trial among the historical trials. The method applies to either binary outcomes or survival to a particular time that is computed from censored survival data. We compute a 95% confidence interval for the predicted result and validate its coverage using simulation. To summarize the additional uncertainty from using a predicted instead of true result for the estimated treatment effect, we compute its multiplier of standard error. Software is available for download.

KW - Principal stratification

KW - Randomized trials

KW - Reproducibility

UR - http://www.scopus.com/inward/record.url?scp=84858865962&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858865962&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2011.01646.x

DO - 10.1111/j.1541-0420.2011.01646.x

M3 - Article

C2 - 21838732

AN - SCOPUS:84858865962

VL - 68

SP - 248

EP - 257

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 1

ER -