|
Activity Number:
|
495
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #310114 |
|
Title:
|
Group-Sequential Analysis of Correlated Binary Data on Repeated Venous Accessions
|
|
Author(s):
|
John Evans*+ and Brian P. Johnson and So Jung Imm
|
|
Companies:
|
Boston Scientific Corporation and Boston Scientific Corporation and Boston Scientific Corporation
|
|
Address:
|
100 Boston Scientific Way, Marlboro, MA, 01752-1242,
|
|
Keywords:
|
Group-sequential design ; Repeated measures ; Correlated binary data
|
|
Abstract:
|
Evans et al. (2006) gave sample sizes and analyses for correlated binary data across time in a venous access study to compare implantable ports. However, the study also required an interim analysis. Group-sequential designs and analyses for uncorrelated binary data are not appropriate for correlated binary data. Incorrect conclusions may be drawn when comparing proportions of complications. We will present group-sequential methods for correlated binary repeated measures in a venous access study. As for any group-sequential design, sample size is slightly higher than for doing just a final analysis, to allow for the interim analysis to be done, whilst preserving the overall significance level. We will compare results from interim and final analyses under the usual assumption of independence of accessions, with those obtained when a positive correlation is assumed among accessions.
|
