Opioid Overdoses among Medicaid Beneficiaries: Predictors, Outcomes, and State Policy Effects - Reducing opioid overdose is an urgent national goal that needs to be supported with the best available understanding of the clinical, treatment, and health system factors that predict overdoses, and points of intervention for their reduction. For individuals surviving medically treated overdoses, there is a need to better understand the factors affecting risk of fatal overdose, including effects of post-overdose treatment, in order to reduce the death toll in this high-risk population. To inform systemic intervention, there is an urgent need for evidence on the role of state policies in reducing risk and improving outcomes. These issues are particularly vital in the Medicaid population, which contributes disproportionately to the national overdose toll. To build evidence on these vital issues, we will use 45-state Medicaid data from 2001-2019, linked to the National Death Index (NDI), supplemented with regularly updated state data from New Jersey and Missouri, to identify the interacting factors contributing to overdose risk and to outcomes for overdose survivors. Under Aim 1, overdose risk will be examined in national incidence cohorts of Medicaid beneficiaries with new episodes of persistent opioid use, and those with new diagnoses of opioid use disorder (OUD). To analyze both non-fatal and fatal overdoses, we will link Medicaid data with the NDI for 50,000 individuals with new persistent use, and 50,000 individuals with new OUD diagnoses, for each year of the Medicaid data, totaling 950,000 in each cohort over the study period. Under Aim 2, we will examine treatment engagement and survival following non-fatal medically treated overdoses, linking Medicaid and NDI data on these overdoses to identify factors affecting post-overdose treatment and outcomes. Under Aim 3, we will examine the impact of an important natural experiment in state policy: the 2017 enactment of major New Jersey legislation. Changes included new limitations on opioid prescribing; use of pain management contracts; and expanded access to medication assisted treatment (MAT) for opioid use disorder. Substantial increases in Medicaid rates for MAT and other OUD treatment were implemented concurrently. Using analytic models similar to those in Aims 1 and 2, we will examine changes in overdose risk among individuals with new long- term opioid use, and among those with new OUD diagnoses, before and after the implementation of the new legislation, relative to comparison populations in states that did not implement similar policies. These analyses will advance our fundamental understanding of opioid risks and outcomes in the Medicaid population with its distinctive demographic characteristics and clinical challenges. By grounding intervention policy in a more comprehensive understanding of overdose risk, and the complex intersection of disability, multi-morbidity, and treatment services within which overdose events are embedded, results can change the paradigm of overdose prevention to one that is informed by a better understanding of these events and the individuals experiencing them.
