Activity Number:
|
529
- Contributed Poster Presentations: WNAR
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
|
Sponsor:
|
WNAR
|
Abstract #307002
|
|
Title:
|
Causes, Impact, and Methods for Mitigation of Covariate Imbalance by Treatment Arm in Stepped Wedge Designs
|
Author(s):
|
Erin Leister Chaussee* and Diane Fairclough and Debashis Ghosh
|
Companies:
|
Colorado School of Public Health and Colorado School of Public Health and University of Colorado Anschutz Medical Campus
|
Keywords:
|
stepped wedge;
covariate imbalance
|
Abstract:
|
The stepped wedge (SW) design is a cluster randomized design in which all clusters transition from control to treatment at successive time periods. In the cross-sectional SW, new participants are continuously enrolled across the entire study period, and covariate imbalance by treatment arm could be caused systematically through differing covariate distributions by cluster, over time, and within cluster at the time of crossover into the intervention period. This imbalance could affect estimation of the treatment effect if covariates are also associated with the outcome of interest, and we aimed to evaluate the performance of standard SW analysis models under these conditions. In simulation studies and under certain conditions, we observed bias in the treatment effect estimates from a standard SW analysis model when confounder distributions differed by cluster and within cluster at the time of crossover. We provide guidance on methods for addressing these systematic causes of covariate imbalance that could be applied during the study design and analysis phases.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|