The second evaluation report on Vermont’s All-Payer ACO Model (VTAPM) Agreement found that although the pandemic and a cyberattack on the University of Vermont Health System posed unique challenges in 2020, the Medicare ACO initiative continued to reduce spending and utilization in payment year 3 relative to a comparison group. The initiative also continued to see progress toward population health improvement goals.
The ACO model is a five-year (2018-2022) arrangement between Vermont and the Center for Medicare & Medicaid Innovation (CMMI) that allows Medicare to join Medicaid and commercial insurers to pay differently for healthcare. The goal is to test whether scaling an ACO model across all major payers in the state would incentivize broad care delivery transformation and ultimately reduce statewide spending and improve population health outcomes.
Under the Model, which initiated in 2017 and went into effect in 2018, CMS provided Vermont flexibility in designing a state-specific, all-payer ACO program. In exchange, the Model State Agreement held that the state is accountable for meeting statewide scale population targets (i.e., Model participation), financial targets, and population health targets.
In 2022, Vermont and CMMI agreed to an extension for one year (2023) with an additional transition year at the option of the state (2024).
The evaluation report, from NORC at the University of Chicago, noted that all but one eligible hospital participated in the Medicaid ACO initiative. Approximately half of eligible hospitals participated in the Medicare ACO initiative—one fewer than in performance year 2. Most critical access hospitals in Vermont’s rural areas opted not to participate in the Medicare ACO initiative, citing the organizational financial reserves required as a barrier to participation.
In addition, the report notes that while the VTAPM was designed to provide an avenue for all payers to participate in payment and delivery system reform through alignment of incentives across payers, it has faced challenges in achieving full participation. “The voluntary nature of payer participation and the state’s limited regulatory ability to influence self-insured employer plans are also challenges to wider commercial payer participation.”
The evaluation team at NORC found that the care coordination infrastructure supported by the Model provided critical support to those most at risk from COVID-19 and helped communities address the public health emergency. Hospital leaders credited the VTAPM as a “catalyst” for increasing collaboration between hospitals and community organizations.
The report notes that the cyberattack on the University of Vermont (UVM) Health Network in October 2020 disrupted the ability of UVM Health Network hospitals and other providers to provide care and bill for services and had a downstream impact on access to care across the state. The evaluation also noted that the pandemic stretched resources and disrupted normal operations for the state, providers, payers, and broader health system partners.
Yet despite the limited scale, NORC found that the Model is achieving gross spending reductions, meaning that some aspects of the Model or contextual factors are still favorable to success.
After accounting for the Medicare shared savings and other pass-through payments, the cumulative net impact of the Model on statewide gross spending across the three full payment years was a significant reduction in Medicare spending of $1,043.58 per beneficiary per year (8.9 percent), or $259.6 million overall. Net shared savings payments to the VTAPM initiative over the first three payment years, taking into account incentive payments to VTAPM providers in the baseline and comparison providers in the baseline and performance periods, totaled approximately $11.1 million. Gross spending decreased during that period by $270.7 million, resulting in a net decrease in Medicare spending of $259.6 million. The net impact through the third quarter of payment year 3was largely similar in magnitude.
“While some state and hospital leaders considered the initial years of the VTAPM to be a period of transition, getting providers on board with the idea of fixed payments and assuming risk, there is still a consensus across interviewees that there has been progress, even while confronting the pandemic,” the report said. “Hospital and community providers credited the VTAPM as a “’catalyst’ for increasing collaboration between hospitals and community organizations and suggested that the Model has increased involvement of different types of providers in care coordination.”
NORC also observed continued progress toward 2022 performance targets for the majority of the Model’s population health and quality of care outcomes. The Model maintained statewide chronic disease prevalence (chronic obstructive pulmonary disease, diabetes, hypertension); increased the Model population’s initiation and engagement for treatment for alcohol and other drug dependence and timely follow-up after ED discharge; and almost halved the percentage of Medicare beneficiaries with diabetes experiencing poor HbA1c control.
As hospitals and providers gain more experience and continue to address the challenges they have faced, the report says, the VTAPM may have a stronger impact on spending, utilization, and quality of care.
NORC said that in future analyses it plans to assess the Medicaid ACO initiative’s reach and impacts, which will cover a larger percentage of attributed beneficiaries than explored in the Medicare analyses. “We will also study ongoing changes in the attributed population’s characteristics and utilization patterns. Additionally, we will quantitatively assess trends in access to specialty and behavioral health services in the ambulatory setting taking into account stakeholders’ perceptions of these trends. All future analysis will continue to consider the lingering effects of the COVID-19 PHE.”