Alcohol use disorders (AUDs) in older adults age 65+ are a substantial and rapidly growing clinical problem. About 40% of older adults are current drinkers, 5.6 million have recently engaged in binge drinking and 1.2 million have a diagnosable AUD. This has been characterized as a “silent epidemic” because AUDs in older adults are vastly underdiagnosed and undertreated in health care. This problem will likely broaden as the number of older adults is expected to increase from 54 million in 2019 to 77 million by 2034. Inadequate access to care for AUD may be a key reason for undertreatment. This is particularly relevant for older adults enrolled in Medicare Advantage (MA) plans, who are subject to rules set by insurers regarding care utilization. These rules include restrictions to receiving care from providers within plan networks. There is a substantial gap in rigorous evidence characterizing the accessibility of providers of AUD care within MA plan networks, and quantifying the effect of inadequate access on older adults' utilization of AUD care. These insights support the long-term goal of our research, which is to guide payers and health systems increase the uptake of care for AUD through rigorous evidence on the effects of provider networks and other managed care approaches. To advance these goals, this R21 project will examine quantify and examine the effects of access to AUD care among older adults enrolled in MA and will accomplish the following three specific aims: 1) describe geographical variation in diagnosis of AUD and unmet treatment need among older adults, 2) descriptively characterize access to providers of services for AUD among older adults, and 3) estimate the relationship between access to care and use of care for AUD. Research in all aims will be conducted using extensive Medicare administrative and claims data, which tracks enrollment in MA and the health service use of Medicare beneficiaries nationally. In Aim 1, geographical variation in AUD diagnosis will be quantified using spatial mixed effects regression models. Exploratory analyses in Aim 2 will quantify access to providers of AUD care across several measures by utilizing comprehensive data on provider networks made available through the Robert Wood Johnson Foundation. Multivariable regression models will be applied to identify individual and local area factors associated with greater access to care. In Aim 3, econometric models for health care utilization will be applied to quantify the relationship between use of AUD care and access measures generated in Aim 2. Products from this study will provide timely data-driven insights for health care stakeholders responsible for ensuring individuals' ability to access care for AUD meets adequate standards. Also, insights identifying the dimensions of access to care that most strongly influence utilization will help stakeholders prioritize approaches to increase uptake of care for AUD. Finally, the current project will genera...