- Alongside Medicare and private payers, states are making the switch to value-based reimbursement, but states and independent researchers have yet to demonstrate the impact of Medicaid alternative payment models on healthcare costs and patient outcomes, a recent Deloitte analysis of 45 models across 28 states revealed.
“Although many state initiatives are underway, relatively few have been evaluated for their impact on total cost of care or health outcomes,” wrote researchers. “One reason may be that many initiatives are relatively new and there has not been sufficient time to observe full program effects. The cost and complexity of conducting formal evaluations also may be a limiting factor.”
The literature review of Medicaid alternative payment models uncovered that state value-based reimbursement initiatives are primarily centered on patient-center medical homes (PCMHs), Medicaid Home Health models, accountable care organizations (ACOs), and bundled payments.
However, evidence was limited on how effective the four Medicaid alternative payment models were with reducing healthcare costs and improving care quality. A recent initiative to catalog alternative payment model evaluations from across payers showed that less than 5 percent of publicly-available evaluations were for state Medicaid programs.
In addition, studies of Medicaid alternative payment models in managed care was limited to case studies, revealing challenges with assessing the aggregate impact of managed care efforts to implement value-based reimbursement and quality improvements.
Deloitte also found that what little data was available on Medicaid alternative payment models was primarily focused on measuring operational success versus quantifying cost reductions.
Medicaid program assessments largely detailed state and/or provider efforts with achieving implementation milestones, such as the percentage of providers using EHRs, the number of community health workers hired, and the number of quality metrics reported.
Data on alternative payment model outcomes tended to be self-reported or anecdotal. Information on provider impact was also usually at the aggregate level with minimal evaluations pertaining to performance variation across participants or among individual providers.
Researchers also reported that few Medicaid alternative payment model evaluations attempted to quantify savings or total care costs. They attributed the absence of information to data limitations, including a lack of complete claims or encounter data access and reporting time lags.
In some cases, states conducted internal analyses to determine cost savings. However, the data could not be externally validated.
While researchers encountered challenges with evaluating Medicaid alternative payment models, the information they did find showed mixed results for the four main models.
Medicaid patient-centered medical homes
The most popular Medicaid value-based care initiative was the patient-centered medical home, with 29 states running at least one in 2016 and 13 planning to develop or expand existing PCMHs in 2017. Despite its popularity, researchers determined that “it is not yet clear whether PCMH initiatives’ operational success will translate into improved health outcomes and reduced costs for patients.”
The majority of reports assessing state and federal PCMH models focused on process measures, rather than quantifiable measures. Out of the few quantitative evaluations, most found minimal measurable effects on claim-based measures of care quality, utilization, and care costs.
For example, a CMS evaluation of the multi-payer Advanced Primary Care Practice demonstration found limited and inconsistent evidence that the PCMH model improved Medicaid outcomes. The assessment showed:
• None of the eight participating states reported substantially lower Medicaid spending growth rates
• Pennsylvania decreased all-cause readmissions and emergency department use for adult beneficiaries
• Michigan reduced all-cause readmissions, but increased specialist visits for children
• North Carolina dropped emergency department use for children
• Only New York and Minnesota saw a boost in primary care visits for Medicaid beneficiaries
However, evaluations of Vermont’s Blueprint for Health and the Community Care of North Carolina PCMHs revealed cost savings. Vermont’s model saved about $482 per participant per year and the model in North Carolina realized annual savings of $312 per Medicaid beneficiary, net of program administration expenditures.
Studies of the Community Care of North Carolina PCMH also showed that the program significantly dropped readmission rates. Although, it did not affect emergency department use.
Researchers did not link Vermont’s Blueprint for Health to care quality improvements for program participants versus a comparison group, except on one diabetes measure.
Medicaid Health Home models
In terms of Medicaid Health Home models, researchers reported that 21 states and Washington DC operated 32 models as of May 2017. The Medicaid alternative payment models borrowed elements from the PCMH and required physician and behavioral health integration. The models also incorporated social and community supports.
Generally, no quantitative analyses on patient and cost outcomes were available, but an HHS evaluation of the Medicaid Health Home model showed that providers and health systems reported improved care coordination, behavioral and physical health integration, and patient engagement.
The participating providers also agreed that the model was “appropriate and feasible to provide targeted care for high-need populations.”
State-level assessments also revealed improved cost and patient outcomes for Missouri’s and Ohio’s Health Home models. Missouri’s primary care model decreased hospital admissions by 5.9 percent and emergency department use by 9.7 percent per 1,000 enrollees, accounting for $5.7 million in cost savings.
The state’s mental health model also realized significant admissions and emergency department use reductions, totaling $2.9 million in savings.
Ohio’s specialty Health Home model increased per-member-per-month spending by $561 in patients with serious mental illness. However, providers reported care coordination and satisfaction improvements and most of the cost increases stemmed from additional care access and pharmacy expenditures.
Medicaid accountable care organizations
The third most popular Medicaid alternative payment model was the ACO, with 10 states operating the model and 13 others looking to implement a Medicaid ACO.
The literature review of existing Medicaid ACOs uncovered that the low implementation rate and relative newness of the alternative payment model resulted in few quantitative evaluations. Most program assessments centered on measuring implementation process versus cost savings or patient outcomes.
Additionally, the review indicated that Medicaid ACOs were maturing slower than expected as evident from studies on two of the oldest Medicaid ACO initiatives. Evaluations of Oregon’s and Colorado’s Medicaid ACO programs reported:
• Colorado’s Department of Health found that the Medicaid ACO model reduced costs by an average of $60 per-member-per-month for adults and $20 per-member-per-month for children, but the cost savings significantly varied by region and the program did not demonstrate improvement in performance, quality, or access measures
• Oregon’s Coordinated Care Organizations program, which operates ACO-like Managed Care Organizations, decreased Medicaid spending growth below the 3.4 percent per year target
• No statistically significant differences between the two models’ effect on performance measures and Medicaid spending, according to an independent study
Medicaid bundled payment models
In 2016, just three states implemented Medicaid bundled payment models and another four announced their intentions to develop the episodic alternative payment model for their program.
Out of the three states engaging with Medicaid bundled payments, only Arkansas published results for its Arkansas Payment Improvement Initiative. The report showed the following impact of Medicaid bundled payments on cost and patient outcomes:
• Perinatal bundled payments did not significantly impact screening rates for pregnant beneficiaries, but the rate of cesarean sections and length of inpatient stays for C-section births modestly declined
• Knee and hip replacement bundled payments improved the 30-day wound infection rate for beneficiaries, but increased the rate of post-operation complications
• ADHD disorder bundled payments boosted the average number of behavioral health visits and reduced the costs of treating patients with the disorder by 15 percent
While the Medicaid bundled payment initiative in Arkansas has yet to show consistent cost savings, Deloitte noted that states may also face challenges with implementing the alternative payment model because of the Medicaid population.
“Much of the bundled payments evidence comes from initiatives targeting conditions and procedures that are high-cost drivers for the Medicare population (eg, joint replacements),” the report stated. “However, there may be insufficient volume in the Medicaid population to achieve savings through those bundles.”
Some states, like Arkansas, are developing bundled payments around common procedures among Medicaid beneficiaries (eg, asthma and maternity bundles), but limited evidence may challenge states as they define care episodes and calculate appropriate bundle rates.
Advice for providers considering a Medicaid alternative payment model
With mixed results for Medicaid alternative payment models, Deloitte suggested that models should add two-sided financial risk to boost provider participation and advance value-based Medicaid reimbursement.
Aligning Medicaid value-based reimbursement models with MACRA’s Advanced Alternative Payment Model requirements would encourage providers to join the programs. CMS plans to include other payer alternative payment model participation to the Quality Payment Program track by the 2019 performance period.
Eligible clinicians could earn a 5 percent incentive payment for sufficiently participating in qualifying Medicaid alternative payment model that includes certified EHR use, value-based payments, and more than nominal financial risk.
Deloitte found that most Medicaid alternative payment models would need to include downside financial risk to qualify as Advanced Alternative Payment Models.
Prior to Medicaid alternative payment model implementation, researchers also advised providers to conduct an assessment to understand how the different models could impact their revenue, relationships with other providers, patient engagement, and their competitive position.
They should also evaluate their capacity to participate in alternative payment models, including their technological and analytic infrastructure and financial risk acceptance.
“The ability to aggregate, analyze, and share data can be essential to identify high-needs, high-risk patients, coordinate care, report on quality measures, and track spending,” stated the report. “Organizations may need to invest in new technology, staff, or training to successfully participate in new APM arrangements.”