How Do We Measure and Account for Treatment Compliance in Community-Based, Multi-Layered Intervention Trials

• Activity Number: 670
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Social Statistics Section
• Author(s): Pan Wu*+
• Companies: University of Rochester
• Keywords: Causal Effect ; Noncompliance ; Distribution-Free Models ; Randomized Trials ; Selection Bias

Abstract

The ITT analysis is the gold standard for examination of treatment effects for clinical trials. However, in community-based, multi-layer treatment studies, compliance is a more complex issue. For example, in child resilience studies, the intervention involved partnership with parents to teach children a set of skills to strengthen emotion self-regulation, adaptive social behavior and classroom conduct. However, parent participation was quite low in this study, despite the near perfect compliance for the children. Additionally, it is not possible to account for parent participation by simply including it as a covariate variable, since parent involvement was not required for the children in the control condition. In this talk, I will discuss a new class of distribution-free models to address the causal treatment effect. The new approach not only overcomes computational problems such as slow convergence rates for existing methods, but also applies to a wider class of data distributions. The proposed methods can be applied to community-based, multi-layered randomized controlled trials, and also to semi-randomized trials such as emerging adaptive intervention study designs.