Evaluating and Optimizing Care for Opioid Use Disorder using a Structured Data-Science Approach

PROJECT SUMMARY/ABSTRACT Opioid use disorder (QUO) is a public health challenge that affects millions of people worldwide. Evidence supporting the efficacy of three FDA-approved medications for OUD (MOUD), including methadone, buprenorphine and naltrexone, is well established. MOUD also constitutes an important intervention in patients who are dually diagnosed with OUD and HIV infection; successful use of MOUD is associated with improved HIV viral suppression. However, most patients with OUD remain untreated. Expanding access to MOUD has been hampered by various challenges in the cascade of OUD care (i.e., diagnosis, treatment initiation and retention, and recovery). Important questions remain about which factors may enhance retention in treatment and reduce both dropout from and cycling on and off MOUD therapy. Furthermore, enhancing knowledge about real-world effectiveness and guiding optimal use of MOUD in the clinical care of patients is also of critical importance. This proposal will use data from the Veterans Health Administration, the largest integrated healthcare system in the United States, along with rigorous analytic techniques, to address this urgent knowledge gap centering around OUD care. In Aim 1 (K99), I will characterize dynamic cascades of care among patients with OUD and among patients with both OUD and HIV, and identify predictors of treatment engagement, disengagement, and re-entry into treatment. In Aim 2 (K99/R00), I will evaluate real-world effectiveness of clinical management strategies on OUD and HIV, including the effect of buprenorphine prescription in primary care settings versus in substance use specialty settings on OUD outcomes, and the effect of MOUDs on HIV viral suppression among patients with both OUD and HIV. In Aim 3 (R00), I will estimate optimal treatment strategies for MOUD to guide personalized treatment decisions. This proposed research will provide a structured model for evaluation and optimization of OUD care. I received my Ph.D. in Epidemiology from the University of North Carolina Gillings School of Global Public Health, with research focus on HIV/AIDS. I am currently a postdoctoral associate in the Department of Epidemiology of Microbial Diseases at the Yale School of Public Health. The career development plan, which aligns with the research proposal and a multidisciplinary mentorship team, outlines a comprehensive strategy for acquiring expert knowledge and analytical skills for my research career. Specifically, I will receive training in: 1) science in substance use research and pharmacoepidemiology, 2) expert knowledge on the intersection of OUD and HIV, 3) skills in dynamic modelling and stochastic processes, and 4) integration of machine learning methods and causal inference theory. Successful completion of this training plan will provide me with skills and knowledge necessary to become an independent investigator and epidemiologist studying infectious diseases and substance use with advanced causal inference and big data analytics.