Characterizing Complex OUD Care Trajectories and Outcomes following Acute Service Utilization: A Population-Based Data Linkage Study - Opioid use disorder (OUD) is a leading cause of morbidity and mortality across the U.S, with wide ranging effects across population subgroups. Medications for opioid use disorder (MOUD) substantially improve health and reduce overdose risk, but engagement and retention in MOUD remains low, with the most high-risk patients cycling in and out of care. Hospital systems frequently interact with high-risk patients with OUD, and are increasingly piloting innovations to improve OUD treatment engagement. However, little is known about the influence of such initiatives on long-term patient care trajectories and outcomes, or how they impact different types of patient subgroups. Studying these trajectories is critical but challenging as disjointed healthcare data complicate tracking outcomes of individuals who move across or outside health systems. Moreover, traditional epidemiologic methods are not set up to characterize complex, heterogenous, and non-linear patterns of OUD care trajectories across multiple timepoints and settings. The current proposal aims to apply novel research approaches to better characterize real-world OUD treatment trajectories following opioid-related hospital encounters, how they differ across patient subgroups, and measure their relationship to hospitalization, overdose and mortality outcomes. To do this, we will focus on patients who experience opioid-related hospital encounters between 2021-2024 at any NYC hospital, with an in-depth focus on patients receiving care at Health + Hospitals, the largest safety-net health system in the U.S. We will partner with public NYC agencies to link multiple administrative databases, including all-payer hospital discharge records, electronic health records, Medicaid claims, neighborhood census data, and mortality records at the patient-level. We will then apply State Sequence Analysis - a novel methodology that uses machine learning to identify clusters of patients that share similar trajectories over time - to: (1) Characterize real-world trajectories of OUD treatment in the year following hospital encounters among treatment-naïve individuals, and determine how care pathways differ based on hospital-centered interventions to initiate MOUD; (2) Assess how OUD care trajectories differ across different patient subgroups, using individual and neighborhood characteristics to identify gaps in continuity of care; and (3) Examine relationships between longitudinal OUD treatment trajectories and subsequent hospital encounters, overdose, and mortality. This methodologically and conceptually innovative proposal will advance knowledge on patterns and gaps in OUD treatment among a high-risk population with poor access to care, and generate evidence on the influence of hospital interventions and treatment modalities on hospitalization and mortality. The unique collaboration between academic researchers and health system and government leaders will ensure translation of findings to actionable treatment and policy interventions, thus advancing NIDA’s goal to improve delivery of high quality, effective services for prevention and treatment of substance use disorders and related outcomes.
