Abstract Details
Activity Number:
|
234
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 4, 2014 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #312650
|
View Presentation
|
Title:
|
A Predictive Modeling Approach to Identify Clinical Important Treatment Difference in a Subgroup: Fraternal Twins Method
|
Author(s):
|
Max Kuhn*+ and Birol Emir and Ed Whalen
|
Companies:
|
Pfizer and Pfizer and Pfizer
|
Keywords:
|
Data Mining ;
Predictive Modeling ;
Subgroup
|
Abstract:
|
As Good (1983) stated post-hoc analysis of the data is "dangerous, useful, and often done". However, within the ultimate framework of hypothesis/idea generation, it may become a "very fruitful" exercise. In this project we assumed (without loss of any generalizability) that the dataset contain p-(continuous) predictors, and a continuous outcome variable from two treatment groups (n from the drug first and m from the placebo). The algorithm is as follows: similar to Hodges-Lehmann (1973) we calculate all pairwise differences between the drug and placebo on the outcome variable and the predictors. For a clinically important cutoff C, we consider only the outcome differences > C, and use those in a predictive method such as CART. Cross Validation is used both for issues dealing with over fitting and dependency. We present an application on a clinical data and simulations are conducted to compare this method with Virtual Twins (2011), SIDES (2011) and ARF (2004).
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2014 program
|
2014 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Professional Development program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.