B1970
Title: Modeling and estimating the effects of cumulative just-in-time treatments using data from micro-randomized trials
Authors: Daniel Almirall - University of Michigan (United States) [presenting]
Inbal Nahum-Shani - University of Michigan (United States)
Susan Murphy - Harvard University (United States)
Abstract: Mobile health interventions aim to provide individualized support, whenever, and wherever it is needed. This includes the provision of therapeutic support in (near) real-time, as well as the provision of prompts that support the engagement of users in the mobile health application. Micro-randomized trials (MRT) are used to address scientific questions concerning the construction of mobile health applications of this type. In an MRT, participants may be randomized hundreds of times over the course of the study. Often, MRTs are designed to learn whether--and if so, when and based on what self-report or passive/sensor data--to intervene with a prompt, suggestion, or some other form of just-in-time treatment. An important set of scientific questions for behavioral intervention scientists involves the cumulative effect of just-in-time treatments. These questions are designed to better understand the dose-response effect of accumulated just-in-time treatments, including how these effects vary over the study. A definition for ``cumulative just-in-time effects'' is introduced in terms of potential outcomes, which is suitable for the data arising from an MRT. It develops a regression approach for estimating these effects. The approach is illustrated using data from an MRT designed to inform the development of a just-in-time adaptive intervention for promoting real-time, real-world engagement in evidence-based self-regulatory smoking cessation strategies.
