Adaptive Messaging to Support Depression Self-Management

点击翻译按钮获取中文摘要

Summary Automated messaging has significant potential to extend access to guidance and support for those managing common mental health conditions like depression. However, benefits of automated messaging could be improved through building capacity to automatically adapt messaging approaches to users’ needs. The PI’s career goal is to establish scalable digital mental health interventions that deliver individualized communication to maximize users’ engagement in effectively managing mental health concerns. This proposal outlines training and research plans to achieve this goal, culminating in a successful R01 proposal. The training plan takes full advantage of the strong institutional support and environment at Northwestern University's Feinberg School of Medicine and the Center for Behavioral Intervention Technologies, which is led by the primary mentor for this proposal. Training goals necessary to the PI’s career goal include growth in (1) user-centered design, (2) adaptive interventions, and (3) clinical science methodologies. This proposal builds on the PI’s background in communication science, which has focused on effective message design and social support provision. In this K01, the PI will aim to develop a model for delivering automated, individualized self-management support for depression, emphasizing messaging that can accommodate fluctuations in users’ motivations and abilities to carry out cognitive and behavioral self-management strategies. This will be achieved through the following specific aims: (1) conduct user-centered design activities and usability testing to design and refine a messaging system; (2) integrate a reinforcement learning algorithm into the system that adapts content to users and their contexts; and (3) pilot procedures and analyses for a randomized controlled trial and obtain preliminary evidence of the intervention’s effectiveness and mechanisms of action. For Aim 3, I will test hypotheses that messaging interventions will result in greater reduction in depression than a waitlist control group, and that adaptive messaging will reduce depression relative to non-adaptive messaging by producing greater engagement in self-management. This research is expected to advance the design of messaging interventions for depression, ultimately extending the reach and effectiveness of these interventions. Skills acquired by carrying out the research proposed in this application will allow the PI to successfully compete for R01 funding in order to conduct a larger effectiveness trial.