Addressing racial disparities in opioid overdose deaths using an open source peer recovery coach training and multimodal mobile health platform

Among people who use opioids (PWUO) some communities bear a disproportionate burden of opioid-related overdose deaths. Health outcomes among PWUO may be worsened by multilevel challenges, including lack of housing, social support, and patient education, which may be partially addressed using scalable opensource solutions (e.g., peer recovery coaches, mobile health, telemedicine). With NIH support, we have validated theory-based, artificial intelligence (AI)-driven texting using natural language processing in a publicly-funded health system to provide real-time responses to patient queries combined with automated texts to enhance identification and address participant unmet social needs (e.g., access to peer support, healthcare), and linkages and retention in buprenorphine treatment per the medical management model (e.g., patient education, self-management). We have also refined an evidence-supported Patient-Centered and Humility (PCH) training for peer recovery coaches (PRCs) that further enhances efforts to address social needs while also including stigma reduction, health habitus, and patient navigation to improve clinical outcomes and social services use among PWUO. This proposal aims to adapt the evidence-supported PCH training to interactive video modules for PRCs, and refine an AI-driven texting tool to reinforce core PCH principles. The 3-month pilot RCT will test AI-driven PCH-enhanced texting tool + PCH-trained PRC providing telephonebased services coordination (multimodal intervention arm-1) vs. AI-driven PCH-enhanced texting tool (intervention arm-2) vs. treatment as usual / informational pamphlets only (control arm) among emergency department enrolled PWUO (N=150) to estimate rates of PRC-PWUO contact, AI-driven texting use among PWUO, assess potential for pre-/post-intervention impact on reduction of felt and/or enacted stigma, clinical outcomes (i.e., time to initial receipt of buprenorphine), and social services use at 1- and 3-months to obtain intervention effect sizes, and estimate the associated cost of implementing and sustaining the multimodal intervention (e.g., cost-per OUD treatment days). The RE-AIM framework will inform implementation of this multimodal intervention that has high public health impact on complementing the HHS Overdose Prevention Strategy (i.e., harm reduction, OUD treatment, and recovery support) in a manner that is scalable and sustainable, and using open-source solutions. Our team is uniquely positioned to obtain data that will inform a subsequent large-scale efficacy trial evaluating our proposed multimodal intervention to increase access to buprenorphine and social services among PWUO.