Longitudinal Modeling of Dynamic Treatment Regimes in the Analysis of Sequentially Randomized Trials

• Activity Number: 17
• Type: Topic Contributed
• Date/Time: Sunday, August 4, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Xi Lu*+ and Daniel Almirall
• Companies: University of Michigan and University of Michigan
• Keywords: adaptive treatment strategies ; GEE ; repeated measures

Abstract

The management of many health disorders, such as depression, diabetes, or obesity often entails a sequential, individualized approach whereby treatment is adapted over time in response to the evolving status and needs of the individual. This decision making can be guided by dynamic treatment regime (DTR), a sequence of individually-tailored decision rules that specify whether, how, or when to alter the intensity, dosage or type of treatment(s) at critical clinical decision points in the course of care. Sequential multiple assignment randomized trials (SMART) were developed explicitly to develop high-quality DTRs. This talk presents a methodology to analyze longitudinal outcomes using SMART data. The methodology can be used to compare mean trajectories among the DTRs embedded in a SMART. Specifically, we describe the use of a weighting-and-replication estimator that exploits the fact that some SMART participants are consistent with multiple embedded DTRs. We also discuss modeling considerations, including constraints on how to model the effect of time. As an example, we present a data analysis of a SMART to develop a DTR for minimally verbal children with autism spectrum disorder.