Online Program Home
  My Program

Abstract Details

Activity Number: 358 - SPEED: Statistics in Epidemiology
Type: Contributed
Date/Time: Tuesday, August 1, 2017 : 10:30 AM to 11:15 AM
Sponsor: Section on Statistics in Epidemiology
Abstract #325117
Title: Immortal Time Bias in Observational Studies of Time-To-Event Outcomes: Assessing Effects of Post-Mastectomy Radiation Therapy Using National Cancer DataBase (NCDB) Study
Author(s): Parul Agarwal* and Erin Moshier and Meng Ru and Madhu Mazumdar
Companies: Icahn School of Medicine at Mount Sinai and Icahn School of Medicine at Mount Sinai and Icahn School of Medicine at Mount Sinai and Icahn School of Medicine at Mount Sinai
Keywords: immortal time bias ; cox regression model ; time-dependent ; chemotherapy ; survival ; breast cancer
Abstract:

Use of optimal methods for addressing immortal time bias (ITB) arising from the variation in timing of initiation of radiotherapy is lacking in radiation oncology literature. The objective of the study is to illustrate ITB by using an oncology outcomes database and quantify through simulations the magnitude and direction of ITB when different analytic techniques are used. A cohort of 9,300 women who received neoadjuvant chemotherapy and underwent mastectomy with pathologically positive lymph nodes were accrued from the National Cancer Database (2004-2008). Multivariable Cox regression models comparing overall survival in patients receiving post-mastectomy radiation therapy with those who did not receive radiation showed a borderline significant treatment effect (HR: 0.94; 95% CI 0.87,1.00). Time-dependent (TD) and landmark (LM) methods estimated no treatment effect with HR: 0.98; 95%CI 0.91,1.05 and HR range: 0.90,1.25, respectively, with LM time from diagnosis ranging from 6-24 months. Simulation study based recommendation will be presented to reduce effect of ITB indicating that TD exposures need to be included as TD variables in hazard-based analyses.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association