Activity Number:
|
359
- Contributed Poster Presentations: Biopharmacutical Section
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #327212
|
|
Title:
|
Nonparametric Survival Analysis with Delayed Treatment Effect
|
Author(s):
|
Kijoeng Nam* and Nicholas Henderson and Dai Feng
|
Companies:
|
Merck and Johns Hopkins University and Merck
|
Keywords:
|
|
Abstract:
|
Clinical trials involving novel immuno-oncology therapies exhibit survival profiles which violate the proportional hazards assumption. In such settings, the treatment of interest may have a delayed effect where the survival curves in the two treatment arms may largely overlap or cross before the two curves separate. To flexibly model such scenarios, we describe a constrained non-parametric approach which allows the survival functions to either not cross or cross at a single point without making any additional assumptions about how the survival curves in the two treatment arms are related. By utilizing constrained splines and a survival-profile dependent step function to model the log-cumulative hazard functions, our method provides interpretable measures of treatment while retaining the flexibility to account for survival profiles commonly seen for treatments with delayed effects. We demonstrate the use and effectiveness of our approach with a series of simulation studies.
|
Authors who are presenting talks have a * after their name.