Improving diabetes outcomes and health disparities by addressing unmet resource needs- A Sequential Multiple Assignment, Randomized Trial - PROJECT SUMMARY Multiple adverse social risks (e.g. food and/or housing insecurity, transportation challenges, social isolation) and out-of-pocket, disease-related expenses are key reasons 1/3 of people with diabetes have high A1cs. Through our CareAvenue mHealth intervention (R01DK116715), we enhanced our original intervention of screening for social risks and connecting people with diabetes to resources by adding features for observational learning, autonomy support, action planning, and self-monitoring. We observed low engagement with the intervention and effectiveness for only 5% of participants. Social support interventions (e.g. peer support, financial navigators, social workers) are more effective in addressing unmet social needs and self- care challenges than stand-alone technology tools. However, these strategies are more resource- and labor- intensive. Rather than dispensing the same fixed package of treatment components to all patients, an adaptive approach can conserve resources by initiating autonomy-supportive treatment, and stepping up treatments for those with suboptimal uptake. This R01 builds on the productivity, infrastructure, and investment of our prior work to address key knowledge gaps for uptake of social care assistance to improve disease outcomes. The overarching goal of this application is to identify what type of supportive components are optimal for addressing unmet social needs and diabetes self-care, and which of five adaptive sequences of treatment results in better outcomes. We will recruit 594 people with diabetes who have high A1cs, unmet social needs, and want assistance with their needs. We will use a Sequential Multiple Assignment Randomized Trial (SMART) design with 12-month follow-up to conduct this research. In Aim 1, we will determine which of five adaptive intervention sequences is optimal for reducing A1c compared to social needs app only (usual care): 1) App + peer support; 2) App + social worker; 3) App + technology-supported financial navigation; 4) App + peer support + social worker; 5) App + peer support + technology-supported financial navigation. In Aim 2, determine which of two augmented treatment adaptations is optimal for reducing A1c among non-responders: social worker or technology-supported financial navigation. In exploratory Aim 2A, we will identify patient-level moderators of treatment effect to inform personalized, resource-efficient protocols. In Aim 3, we will estimate cost-effectiveness of five adaptive intervention sequences. We hypothesize that at 12 months, App + Peer Support augmented with technology-supported financial navigation will produce superior outcomes (reduced A1c, uptake of social care resources) relative to the 4 other adaptive interventions. The result of this SMART study will be an optimized, adaptive intervention to improve the health and social wellbeing of people with diabetes by determining the most effective intervention strategies.
