|
Activity Number:
|
482
- Causal Inference and Related Methods
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2018 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
|
Abstract #330758
|
Presentation
|
|
Title:
|
Principal Stratification for Longitudinal Data in Environmental Trials
|
|
Author(s):
|
Joshua Keller* and Roger D Peng
|
|
Companies:
|
Johns Hopkins Bloomberg School of Public Health and Johns Hopkins University
|
|
Keywords:
|
Environmental Trials;
Causal Inference;
Immunology;
Indoor Air Pollution
|
|
Abstract:
|
Randomized trials of indoor air quality interventions have shown promise for identifying approaches for improving indoor air pollution and reducing adverse respiratory symptoms in children. A recent two-city trial targeting mouse allergens compared an active intervention of professional pest management services against education alone. An intent-to-treat analysis found no significant difference in repeated measures of respiratory symptoms between treatment and control. However, the intent-to-treat approach ignores available information about the impact of treatment assignment on actual mouse allergen concentrations. Using a principal stratification approach, we estimate the potential health benefit among subjects who would have a reduction in mouse allergen levels when assigned to treatment and among those who would see no notable change in mouse allergen levels under treatment. This analysis demonstrates the applicability of principal stratification for environmental trials with continuous exposures and repeated observations.
|
Authors who are presenting talks have a * after their name.
