|
Activity Number:
|
383
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Risk Analysis
|
| Abstract - #309236 |
|
Title:
|
Parametric Modeling of Melanoma Prognoses: An Alternative to the Cox Regression Model
|
|
Author(s):
|
Shouluan Ding*+ and Huiyi Lin and Renee Desmond and Seng-jaw Soong+
|
|
Companies:
|
The University of Alabama at Birmingham and The University of Alabama at Birmingham and The University of Alabama at Birmingham and The University of Alabama at Birmingham
|
|
Address:
|
NP 2540, Birmingham, AL, 35242-3300, NP 2540, Director, Division of Biostatistics and Bioinformatics, Birmingham, AL, 35242-3300,
|
|
Keywords:
|
AFT Models ; Melanoma Survival Data ; Extended Generalized Gamma Distribution ; Parametric models
|
|
Abstract:
|
We investigated the accelerated failure time (AFT) parametric survival model as applied to multivariate prognostic modeling of localized melanoma data with six known potential prognostic factors. Using probability plotting, Cox-Snell residual plots, and likelihood-ratio statistic test methods, we demonstrated that an extended generalized gamma (EGG) survival function was the best fitted parametric model applicable to melanoma prognosis. Using this criteria and the relationship of scale and shape parameters of EGG, all showed that the EGG model adequately served as an alternative model to Cox regression model in accurately predicting survival rates in melanoma. The EGG model provided more efficient parameter estimates compared to the Cox model. It also estimated hazard and survival functions for individualized patient prognosis, making it more useful for clinical interpretation.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
