Signature Project

点击翻译按钮获取中文摘要

Signature Project Abstract Community colleges provide a critical pathway for workforce development and socio-economic gain, but this opportunity is mitigated by unmet need for mental health services, particularly for depression and anxiety, and particularly for racial/ethnic minority students. Mental health problems intersect with high rates of food insecurity, housing insecurity and homelessness to create a mutually exacerbating cycle of disability and impairment. A scalable and effective system of care that manages mental health needs in concert with social mental health determinants is sorely needed. The Signature Project aims to optimize the effectiveness of scalable system of care, called STAND, in low income, highly diverse sample at East Los Angeles College, while concurrently advancing the science of personalized mental health care. Greater personalization of tools for clinical decision-making offers more efficient and more effective allocation to treatments that reduce attrition and improve symptomatic and functional outcomes. Personalization includes selecting the appropriate level of care at initial presentation and knowing when a change to level of care is needed. This is the first study to address tools for personalizing care that are suited to the needs of diverse community college students. The goal of the Signature Project is to compare standard of care approaches for treatment triaging and adaptation (based on symptom severity alone) to data-driven algorithms that draw from four overlapping and mutually reinforcing constructs: social determinants of mental health; early adversity and life stress; predisposing, enabling and need influences upon health services use; and comprehensive mental health status. Static features at baseline as well as trajectories over time will be used to allocate to initial level of care (ranging from self-guided online prevention, to coach-guided online therapy, to clinician delivered care) and inform adaptation of care (e.g., stay the course, increase level of care, re-initiate care) over 40 weeks. ELAC students will be randomized (N=200 per year, total N=1000) to either symptom severity-based decision making or to data- driven decision making. We will evaluate whether the data-driven algorithms improve adherence, symptoms of depression and anxiety, and functioning. The measures and algorithms will be refined annually based on statistical prediction, review by the Methods Core Health Disparities/Cultural Competency group, and observational data from the Healthy Minds Survey results from 10 community colleges across California (including ELAC). The end goals are to optimize the effectiveness of the STAND program for ELAC students, while creating templates for personalization to be sustained and spread to other community colleges, and generating data that will inform cost-effectiveness and return-on-investment projections for implementation of STAND region or state-wide.