ABSTRACT From April 2020, the start of the COVID-19 pandemic in the US, to April 2021, there were more than 100,000 drug overdose deaths in the US. It is estimated that the drug overdose epidemic currently costs the US about $1 trillion each year. Over the past 5-7 years, drug overdoses involving multiple drugs (polydrug) has increased especially rapidly. More than 60% of all drug overdose deaths in the US now involve multiple drugs, especially illicit fentanyl mixed with prescription opioids, cocaine, benzodiazepines, and stimulants. In North Carolina (NC), the state where the proposed study focuses, more than 70% of overdose deaths involve polydrug overdoses, including 53% which involve at least one opioid and 19% which involve multiple non- opioids. Polydrug overdoses are also increasing disproportionately among racial/ethnic minority Americans. Systems like healthcare, insurers, and corrections, which have the potential to implement large prevention initiatives, lack the linked data to establish benchmarks of polydrug overdose incidence in their populations and define risk and protective factors. Many people who die from polydrug overdose have had recent contact with healthcare, insurance, or correctional systems. In fact, drug overdose death rates from individuals on Medicaid and those released from correctional settings are much higher than the general population. Further, the role of prescription opioids for acute, post-surgical, or chronic pain in leading to polydrug use and overdose is not understood. It is important to understand pathways to polydrug use and barriers to substance use disorder treatment through the lens of lived experience of persons who use drugs and treatment providers. Therefore, in this mixed-methods study, we will define polydrug overdose rates and trends in four large populations representing publicly and privately insured individuals, health system patients, and those released from incarceration over a 16-year period; identify demographic, clinical, and short- and long-term care access patterns and factors that predict risk of fatal polydrug overdoses in each of the four populations; identify prescription opioid pain management trajectories in longitudinal cohorts of patients with acute, post-surgical, and chronic pain that are associated with increased risk of fatal polydrug overdoses; and elucidate the pathways to polydrug use and overdose people with lived experience and understand barriers to effective substance use disorder treatment and linkage to such treatment from providers and harm reduction professionals. We will also examine racial disparities and racial algorithmic bias in prediction models and develop tools that may aid clinicians and corrections officers to identify people at high risk for polydrug overdose deaths and connect them with appropriate treatment options. This work will leverage data resources and linkages developed from our prior and ongoing studies, thereby maximizing resources, and greatly advancing our ability to develop evidence-based interventions through health systems, insurance providers, and corrections in North Carolina and the United States.

PROJECT NARRATIVE This project will define polydrug overdose mortality rates for four key target populations of patients either engaged in health care, privately or publicly insured, or recently released from prison; will define clinical predictors of polydrug overdose mortality risk; will identify prescription opioid pain management trajectories associated with polydrug overdose in longitudinal cohorts of pain patients; and identify pathways to polydrug use and overdose.