A Nonparametric Approach for Comparing Survival Distributions with Propensity Score Adjustment

• Activity Number: 26
• Type: Contributed
• Date/Time: Sunday, August 3, 2008 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Nandita Mitra*+ and Katrina Armstrong and Glenn Heller and Venkatraman E. Seshan
• Companies: University of Pennsylvania and University of Pennsylvania and Memorial Sloan-Kettering Cancer Center and Columbia University
• Keywords: propensity score ; survival data ; nonparametric test statistic ; Medicare

Abstract

We propose a nonparametric test for the equality of the conditional hazard function between comparison groups [Heller and Venkatraman, 2004] that allows for the adjustment of concomitant variables via the estimated propensity score. This conditional logrank test does not require the assumption of proportional hazards, a parametric specification of the relative risk function, or randomization of group assignment. The methodology is demonstrated in a large observational study of 76,286 Medicare enrolled prostate cancer patients that compares the survival benefit of men who received prostatectomy versus those who did not. Numerous covariates such as race, age, socio-economic status, and comorbidities were adjusted for via the propensity score.