Commonality Analysis for Survival Data, with an Application to Data from Breast Cancer Patients with Newly Diagnosed Brain Metastases

• Activity Number: 38
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Binglin Yue and Haitao Chu and Paul Sperduto and Xianghua Luo*+
• Companies: University of Minnesota and University of Minnesota and University of Minnesota at Gamma Knife and University of Minnesota
• Keywords: Coefficients of determination ; Commonality analysis ; Generalized gamma regression ; Survival analysis

Abstract

The commonality analysis is widely used in social and behavioral research. It is a method for determining the amount of variability in a continuous outcome variable explained by multiple predictors, individually or jointly. However, the original commonality analysis method was applicable for only continuous outcome variable. In this paper, we propose to use commonality analysis for censored survival data. Coefficients of determination (R2), derived from generalized gamma regressions are used to quantify percentage of variability explained by predictors for log-scaled time to event. We apply the proposed method to data from breast cancer patients with newly diagnosed brain metastases. Important demographic and clinical factors and their contribution to survival prognosis are determined.