Veteran enrollment in Medicare Advantage: Analyses of favorable selection, plan marketing, and utilization

Abstract: Dual enrollment between the Veterans Health Administration (VHA) and Medicare Advantage (MA), the private alternative to Traditional Medicare (TM), poses a unique budgetary issue for the federal government. MA plans are paid even if their Veteran enrollees use VHA care, leading to duplicate spending by the federal government. Over the last decade, the VHA has spent over $78 billion for dual MA-VHA enrollees. With 1.3 million Veterans dually enrolled in MA and VHA benefits — some in Veteran Medicare Advantage Plans (VMAPs) that market to the Veteran population — there are concerns that duplicate spending has increased alongside the growth of MA. To support policy and payment reform efforts to address the duplicate spending for dual MA-VHA enrollees, in three papers, this dissertation investigates selection of VHA enrollees into MA plans, the benefit designs of MA plans and VMAPs, and the effect of VMAP marketing and benefit design changes on use of VHA care. To investigate selection, in the first paper, I examined differences between MA-VHA and TM-VHA enrollees from 2016–2019 using VHA survey data and VHA Nosos risk score files, which are used to predict Veterans’ expected VHA costs. Since the VHA Nosos risk score data builds on the CMS Hierarchical Condition Category (HCC) model for adjusting MA plan payments, it allowed me to identify previously undetected favorable selection of healthier, lower-cost VHA enrollees in MA. Compared to TM-VHA enrollees, MA-VHA enrollees had Nosos scores that were 25.6 percentage points (pp) lower in 2019 and $453.79 lower VHA costs, on average between 2016 and 2019, suggesting potential overpayments to plans based on selection alone, not accounting for the mix of MA- vs. VA-financed care.In the second paper, I examined what benefits and cost sharing MA plans offered to VHA enrollees in 2022, contrasting VMAP to other MA plans. I applied a novel approach to identifying VMAP plans, marketing to the Veteran population, in publicly available MA plan directory files, and compared their enrollment and benefit design to other MA plans. I found that compared to other MA plans, VMAPs were more likely to enroll priority group 1 Veterans (facing zero cost sharing for VHA care) and offer coverage for hearing, dental, and vision services, Medicare Part B premium reductions, but exclude coverage for Medicare Part D prescription drugs. This work revealed the heterogeneity in MA plans’ benefit design and the potential for VMAPs to attract those who are likely to use the VHA for care. To illuminate duplicate payment, in the third and final paper, I expanded on the work from paper two to examine VHA enrollees’ utilization in VMAPs and other MA plans between 2019 and 2022. Using a quasi-experimental difference-in-difference (DID) design, I leveraged exogenous changes in MA plans’ marketing and benefit design to analyze the effect of VMAPs on enrollees’ use of VHA outpatient care, compared to other MA plans. This analysis focused on a subset of MA plans that rebranded to a VMAP in 2021. On average, between 2019–2022, enrollees in VMAP plans were 72.3% reliant on VHA outpatient services (using more VHA-paid care than Medicare-paid services) while other MA enrollees were 32.2% reliant on the VHA. Following VMAP rebranding in 2021, in VMAP plans, VHA outpatient reliance increased by 2.29 pp (95% CI 1.50, 3.07; p<0.001) in unadjusted models and 1.27 pp (95% CI 0.55, 1.99; p=0.001) in adjusted models. Results were consistent among the subgroup of VMAPs that exclusively changed their plan name and marketing, but I observed even larger effect sizes on increased VHA reliance among the subgroup of VMAPs that added new supplemental benefits as part of their rebranding efforts. These findings support the work in paper two, underscoring VMAPs ability to attract highly VHA reliant Veterans and suggesting that the scale of duplicate spending may be more prominent in VMAPs compared to other MA plans. These results present several novel contributions including the potential factors for favorable selection of VHA enrollees, the identification of nascent VMAPs marketing to Veterans, and the impacts of VMAP marketing and benefit design on selection and duplicative payment. These results can be used by CMS and VHA leadership to inform legislation and payment reform options to address the government’s duplicate spending for dual MA-VHA enrollees.