Activity Number:
|
487
- Topics in Clinical Trials - II
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #313545
|
|
Title:
|
Adjusted Survival Comparisons at Fixed Points in Time
|
Author(s):
|
Kyle Rudser* and John Connett
|
Companies:
|
University of Minnesota and University of Minnesota
|
Keywords:
|
survival analysis;
adjusted;
confounding;
censoring;
nonparametric
|
Abstract:
|
The primary objective is to compare differences in survival curves between groups while accounting for covariates at a fixed point in time. The commonly used Cox Proportional Hazards model could be used, though has several shortcomings, including assumptions to be met for it to perform optimally. Recently developed methods utilizing pseudo-values and separately forming adjusted contrasts take a step towards a non-parametric approach to estimate adjusted differences in survival at a fixed point in time. We present simulations to directly compare estimation and inference for adjusted differences in survival curves between the Cox Proportional Hazards model and alternatives with less stringent assumptions across a variety of scenarios. While the Cox Proportional Hazards model predictably performs better on data that follow the proportional hazards assumption, alternative approaches were less biased and had lower MSE overall on data that did not follow such assumption. An illustrative example in the context of lung transplant patients with and without Alpha-1 antitrypsin deficiency is also presented.
|
Authors who are presenting talks have a * after their name.