March 2012
Volume 53, Issue 14
Free
ARVO Annual Meeting Abstract  |   March 2012
Multi-Dimensional Testing to deal with Multiplicity in Clinical Trials
Author Affiliations & Notes
  • Swetha Surabhi
    Biostatistics,
    Statistics & Data Corporation, Tempe, Arizona
  • Dale Usner
    Biostatistics,
    Statistics & Data Corporation, Tempe, Arizona
  • Kathryn S. Kennedy
    Biostatistics,
    Statistics & Data Corporation, Tempe, Arizona
  • Dale J. Kennedy
    Statistics & Data Corporation, Tempe, Arizona
  • Richard Abelson
    Statistics and Data Corporation, Tempe, Arizona
  • Footnotes
    Commercial Relationships  Swetha Surabhi, SDC (E); Dale Usner, SDC (E); Kathryn S. Kennedy, SDC (E); Dale J. Kennedy, SDC (E); Richard Abelson, SDC (E)
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science March 2012, Vol.53, 4014. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Swetha Surabhi, Dale Usner, Kathryn S. Kennedy, Dale J. Kennedy, Richard Abelson; Multi-Dimensional Testing to deal with Multiplicity in Clinical Trials. Invest. Ophthalmol. Vis. Sci. 2012;53(14):4014.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: : Ophthalmic clinical trials often involve multiple comparisons of different treatment regimes. For example, drug combinations are often compared to their components alone, as well as to the vehicle. Such studies require performing multiple tests of significance, and are prone to Type I error. Any method to address this potential problem should fit within the study design, maximize power and minimize false positives. The aim of this work is to compare the power of two methods, fixed sequence testing and multi-dimensional testing. In fixed sequence testing, hypotheses are prospectively ordered and tested, thus efficacy of subsequent hypotheses cannot be claimed if prior hypotheses were not found to be significant. To overcome this drawback, a multi-dimensional testing framework has been developed, where the decision-making process no longer exhibits a simple sequential structure, but is guided by a decision tree with multiple branches that correspond to individual treatment doses.

Methods: : A simulated dataset was generated with a thousand points using a normal distribution where low dose combo and high dose combo are more effective compared to placebo, but only high dose combo is more effective when compared to the individual components. The prospective order of the hypotheses in fixed sequence testing is low dose combo vs. placebo, high dose combo vs. placebo, low dose combo vs. its individual components and high dose combo vs. its individual components. In multi-dimensional testing, the decision tree has two branches where one branch corresponds to low dose combo and the other branch corresponds to high dose combo. The aim is to find the power of the two methods for testing the efficacy of the combo drugs.

Results: : Power can be increased from 4.1% to 81.4% using the multi-dimensional testing. Since low dose combo is not effective compared to its individual components, the efficacy of high dose combo compared to its individual components cannot be claimed using the fixed sequence testing, this can be overcome using multi-dimensional testing.

Conclusions: : The multi-dimensional testing is a useful approach that can be applied to a wide variety of clinical trials to account for multiplicity. The results demonstrate that multi-dimensional testing can provide power advantage over fixed sequence testing for a scenario that can be encountered in clinical trials.

Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology 
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×