Background: The Veterans' Access, Choice, and Accountability Act of 2014 has transformed the way the VHA delivers care, and its reorganization may lead to substantial challenges that ultimately affect the quality of care that Veterans receive, and subsequently, their health outcomes. Specific Aims: Our specific aims are to: 1) examine variation in utilization and access of VHA vs. Community Care (CC) over time (FY15-FY19); 2) develop and test a methodology to compare costs between VHA and CC; and 3) examine use of specialty care, specifically surgery and mental health. Within Aim 1, we will also determine the factors associated with receiving care in VHA vs. CC. For Aim 2, we will compare cost of standardized episodes of care (SEOCs) in VHA and CC for surgical care, and examine whether use of SEOCs leads to reduced cost and utilization. We will compare quality of surgical care and mental health care between VHA and CC, and develop methods to identify “overuse” of low-value surgical/mental health care in VHA vs. CC. Unique Features/Innovations of Project: This study will evaluate how well the Veterans Choice Program is working to increase Veterans' choice about where they get their care and whether this “choice” has helped to improve their access to timely, high-quality care. Through collaboration with key operational partners (the Office of Community Care (OCC), the Partnered Evidence-based Policy Research Center (PEPReC), and the Office of Reporting, Analytics, Performance, Improvement and Deployment (RAPID)), we will provide VHA with critical information on how well VHA's transformation to a purchaser of care is working. Methodology: For Aim 1, we will conduct descriptive analyses of the overall use of VHA and CC over time. We will then run a series of stratified analyses, separately by inpatient and outpatient setting, by category of care and by SEOC and then at the facility level. We will examine patient- and facility-level characteristics that are related to the extent of use of CC using descriptive analyses and then multivariable logistic regression models adjusted for patient and facility characteristics, categorizing facilities into “higher CC use than expected” and “lower CC use than expected.” For Aim 2, we will use a difference-differences (DD) to estimate costs prior to and after the the use of SEOCs for two surgical SEOCs in VHA and CC (Orthopedics and Neurosurgery-Neuropsych). Similar analyses will be conducted for quantity of services (instead of cost). For Aim 3, we will use generalized linear models, adjusted for patient characteristics and types of services received, to evaluate differences in outcomes between patients using VHA vs. CC for surgery and mental health. Once we have a reliable estimate of the differences in quality of care and cost for the same types of services within SEOCs provided in VHA vs. CC, we will derive a method for identifying low-value care and suboptimal allocation of resources. Expected Results: Inform...