|
Activity Number:
|
370
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 2, 2016 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
ENAR
|
|
Abstract #321062
|
|
|
Title:
|
Causal Inference with Partial Interference and Right-Censored Outcomes
|
|
Author(s):
|
Sujatro Chakladar* and Michael Hudgens
|
|
Companies:
|
and The University of North Carolina at Chapel Hill
|
|
Keywords:
|
Causal Inference ;
Interference ;
Right censoring ;
Survival ;
IPW ;
IPCW
|
|
Abstract:
|
Interference arises when the outcome of one individual depends on the treatment status of another individual. Partial interference is a special case of interference where individuals can be partitioned into groups such that no interference occurs between groups but may occur within groups. Tchetgen Tchetgen and VanderWeele (2012) proposed inverse probability weighted (IPW) estimators for different causal effects in the presence of partial interference. In this poster we extend their IPW estimators to the case where the outcomes may be right censored using inverse probability of censoring weights (IPCW). The large sample properties of the IPCW estimators are derived and simulation studies are presented demonstrating the estimators' performance in finite samples. The methods are illustrated using data from a cholera vaccination trial in Matlab, Bangladesh.
|
Authors who are presenting talks have a * after their name.
