QT-Digital Mental Health Engagement Study

Description

This study aims to evaluate the feasibility of testing tailored engagement strategies with SGM DMH consumers using a micro-randomized trial (MRT) design and to create preliminary evidence about which theory-driven strategies most effectively promote engagement with DMH resources. This study will involve delivering brief, theory driven engagement strategies at two key decision points within the flow of MHA’s Screening Program and assess feasibility of integrating these strategies into MHA’s existing web infrastructure and testing them in a naturalistic setting.

The primary objective is to evaluate whether delivering (vs. not delivering) engagement strategies after an online depression screening increases initial and sustained engagement with DMH resources on the MHA website, and to understand which engagement strategy might be most effective. The engagement strategies are grounded in the Health Action Process Approach (HAPA), including immediate behavioral determinants: outcome expectancy, self-efficacy, perceived risk, and barriers and resources.

Eligibility assessment will begin after a user completes the online depression screening (PHQ-9) hosted on the MHA website as part of their Screening Program. Users who complete the online PHQ-9 and optional demographic questions will be considered for inclusion. Those who are aged 14+ and SGM will be randomized at each decision point to either receive or not receive engagement strategies. Participants will encounter two randomization decision points embedded within the usual user flow of the MHA Screening Program.

The first decision point will be the Results page where users are typically shown information about their PHQ-9 score and provided a set of 3-4 targeted Next Steps resources. Here, participants will be randomized with equal probability (0.2) to either receive: 1) no engagement strategy (i.e., the standard Results page as usual; control), or; 2) one of four types of HAPA-based engagement strategies. Participants randomized to receive an engagement strategy will be randomized with equal probability to an engagement strategy targeting a HAPA construct: 1) outcome expectancy; 2) self-efficacy; 3) perceived risk, and; 4) barriers and resources. On the Results page, these engagement strategies will be displayed as inline messages designed to quickly reinforce a HAPA behavioral determinant before users choose their Next Steps. The second decision point will be the targeted Next Steps resource page that participants click on from the Results page. Here, participants will be randomized with equal probability (0.2) to either receive: 1) no engagement strategy (i.e., the standard Next Steps resource page as usual; control), or; 2) one of four types of HAPA-based engagement strategies. Participants randomized to receive an engagement strategy will be randomized with equal probability to an engagement strategy targeting a HAPA construct: 1) outcome expectancy; 2) self-efficacy; 3) perceived risk, and; 4) barriers and resources. On the Next Steps resource page, these HAPA-based engagement strategies will be displayed as inline cards embedded directly within the page layout. These cards function as small UI components (e.g., badges) that visually stand out but do not interrupt user navigation. Each card is delivered seamlessly near the top of the Next Steps resource page content and can be read without being redirected or leaving the page.

Randomizations are independent across decision points. Delivery of engagement strategies occurs passively within the website UI and does not alter the standard navigation flow.

The primary engagement outcome will include participants’ click-through to the targeted Next Steps resource and total time on the targeted Next Steps resource. Secondary engagement outcomes will include the total time on MHA pages and total number of MHA pages visited beyond the Results and targeted Next Steps resource, capped at 30 minutes of inactivity. All engagement outcomes will be captured automatically through MHA’s analytics system.