Download as PDF

Submitted online at https://hcp-lan.org/groups/pbp/financial-benchmarking/fb-comments/

March 7, 2016

Dear members of the HCPLAN Guiding Committee:

The National Association of ACOs (NAACOS) submits the following feedback in response to the Health Care Payment Learning & Action Network (LAN) Financial Benchmarking Draft White Paper, as released February 8, 2016. NAACOS is the largest organization of Medicare Shared Savings Program (MSSP) ACOs, representing approximately 175 MSSP, commercial, Next Generation and Pioneer ACOs. NAACOS is a member-led and member-owned non-profit organization that works on behalf of ACOs across the nation to improve the quality of Medicare delivery, population health and outcomes, and health care cost efficiency. Our members, more than many other healthcare organizations, want to see an effective, patient‐centric care transition process. Our recommendations reflect our expectation and desire to see the MSSP achieve the long-term sustainability necessary to enhance care coordination for Medicare beneficiaries, reduce healthcare costs and improve quality in the Medicare program.

This white paper provides high level principles on financial benchmarks, which are necessary for population‐based payment models to establish spending levels against which these organizations are measured. Establishing sound financial benchmarks is essential to the success of population-based payment models, such as ACOs, and we are pleased to contribute feedback to the draft financial benchmarking white paper, which addresses a number of areas related to our financial benchmark principles as outlined along with our feedback below. We hope the HCPLAN Guiding Committee considers this feedback as you move forward with your work to accelerate implementation of payment models that promote provider accountability for a patient population across the full continuum of care.

NAACOS financial benchmark principles and white paper feedback

Do not tie benchmarks solely to historical performance
Reset financial benchmarks should not be based exclusively on an ACO's historical expenditure data. That approach penalizes ACOs for performing well in the past and forces them to chase increasingly more challenging benchmarks in subsequent agreement periods. Under this flawed methodology, ACOs face difficult decisions about whether to continue participating with payers as ACOs because their success is punished when reset benchmarks are reduced based on lower spending in prior agreement periods. This policy also reduces the incentive for ACOs to invest in efforts that would reduce spending, a result which is detrimental to the ACO as well as the Medicare Trust Funds.

Allow ACOs to retain savings
Payers use financial benchmarks to hold providers accountable for delivering efficient care and should reward successful providers with the savings they generate. The success of population-based payment models depends on providers delivering high-quality, efficient care. Successful ACOs plan and invest heavily in reducing inefficiencies, coordinating care and preventing adverse outcomes. This includes significant start-up costs and considerable investments in health IT, care coordination and prevention efforts. Further, successful ACOs improve quality of care and outcomes, leading to reduced healthcare spending and foregone revenue for ACOs paid based on fee-for-service (FFS). Therefore, ACOs should be rewarded with the maximum amount of savings they generate, which is the difference between their actual spending and their financial benchmark. After achieving savings relative to their benchmark, successful ACO should not be penalized by facing lower future benchmarks as a result of their savings.

This situation is unfair to providers and creates a downward spiral which must be avoided to maintain voluntary participation in population-based payment models.

Base benchmarks, at least partially, on regional data
Financial benchmarks should be comprised of a blend of historical and regional cost data. We support gradually incorporating regional FFS cost data into reset benchmarks, along with a portion of the ACO's historical costs. We do not support transitioning ACOs from benchmarks based on their historical costs toward benchmarks based only on regional FFS costs or on national costs. Rather than exclusively relying on regional cost data, which does not take into account the patient population of a specific ACO, we support blending historical ACO and regional cost data for non-ACO beneficiaries into reset benchmarks. We do not feel it is appropriate to rely exclusively on regional cost data for reset benchmarks as many ACOs have unique patient populations which necessitate continued use of a portion of historical cost data in reset benchmarks.

Do not include ACO-assigned beneficiaries in the regional population
Payers should remove ACO-assigned beneficiaries from the regional population used to calculate regional costs included in benchmarks. Rather than comparing ACOs to themselves, ACO performance should be evaluated relative to FFS. Excluding ACO-assigned beneficiaries allows for a clearer comparison between ACOs and FFS. Should a payer not remove the ACO-assigned beneficiary population, the regional cost data would be skewed by reflecting ACOs' efforts to coordinate care and reduce expenditures for the ACO population. To address ACOs whose remaining regional population falls below a certain threshold (ex. 5,000) after removing the ACO-assigned beneficiaries, we recommend using a modified approach to reach 5,000 beneficiaries. For example, this gap could be bridged by increasing the weight of the counties that have a lower proportion of ACO beneficiaries residing there. Another option would be to expand the regional area to include beneficiaries in adjoining counties. Overall, it is essential that the regional population exclude ACO-assigned beneficiaries for all ACOs in the region.

Base benchmark trending on regional data
During an agreement period, we support benchmark trending that uses a regional adjustment reflective of the actual market in which the ACO operates. Payers should move away from a single national amount toward regional trending factors to align with incorporating regional FFS expenditures in benchmarks.

This approach better captures the cost experience in an ACO's region as well as the health status and socio-economic dynamics of the regional population. Payers should use this approach with initial and reset ACO benchmarks.

Allow benchmarks to reflect changes in patient acuity
ACOs are responsible for the costs of an attributed population, which requires properly risk adjusting to account for patient acuity. For a continuously enrolled population, some ACO models, such as the MSSP, cap risk scores at the ACO's baseline risk and only allow an increase in the risk adjustment based on demographic changes (e.g., the aging of the population), not on changes in the acuity of the population. On the other hand, the MSSP allows reductions in the risk adjustment based on demographic factors or risk scores for the continuously enrolled. Only counting risk scores that work against the ACO for the continuously enrolled population disadvantages ACOs that take on the management of the sickest populations with the greatest medical need. This policy stems from a concern that the ACOs will receive higher payments due to improved documentation and coding or purposeful abuse of coding. We understand this concern but feel it is unfounded and has not been proven in the ACO community. Correct accounting for changes to ACO risk ensures that providers are paid appropriately and will continue to take on the sickest patients. Artificial risk adjustment caps apply a perverse incentive in which those ACOs that meet the goal of improved patient health, reduced costs through coordinated care management, and other long-term strategies are still penalized. These organizations see a decrease in acuity for well-managed patients, which counts against them, while they do not receive credit for caring for patients whose acuity intensifies. Payers should incorporate the full growth in HCC risk scores across all contract years.

Solidify necessary methodologies
Financial benchmarking and many quality measures require patients to be attributed to a provider or organization and necessitate risk adjustment to account for existing patient acuity. Methodologies for risk adjustment and attribution are essential to performance evaluation, and we call on industry leaders including CMS to conduct the exhaustive research and necessary analysis to create scientifically sound and widely accepted patient attribution and risk adjustment methodologies. For too long the healthcare industry has relied on methodologies plagued with flaws, which can unfairly penalize providers and hold the potential to dramatically undermine the credibility of population-based payment models that rely on quality and cost measure evaluation.

Use transparent processes for establishing benchmarks
The methodologies to determine benchmarks must be transparent, including transparency for how the benchmark is set as well as how the ACO's performance is assessed relative to its benchmark. A lack of transparency undermines provider trust in benchmarks, which, given the significance of the benchmark to a population-based payment model, has the potential to stall adoption of these models across the healthcare industry. It is essential that ACOs have confidence in how their benchmark is set and in how they are evaluated against it. Along with being transparent, payers must also be timely with releasing benchmarks, the data used to calculate them and ACO performance assessment.

Allow flexibility and provider choice
We support providing ACOs with flexibility and options related to certain benchmark methodologies, especially as payers transition their approach to benchmarks over time. For instance, transitioning to benchmarks that blend historical and regional expenditure data affects ACOs differently, with some seeing significant and unexpected swings in their reset benchmarks. This situation would present considerable challenges and may force ACOs harmed by such a change to no longer participate with the payer. In addition to the magnitude of a benchmark change, payers must also consider the pace of the transition related to a benchmark methodology change. Too rapid a transition could be a deterrent to ACOs while too slow a transition could also discourage ACOs, particularly those eager to move to a new benchmark based on a revised methodology. To mitigate unexpected benchmark swings and to ease the transitions from benchmark methodologies, we urge payers to provide a glide path with options for ACOs

to decide how to move to a new benchmark methodology. Providing maximum flexibility and choices best serves ACO and payers, both of which benefit from continued participation in population-based payment models.

When feasible, provide benchmarks in advance of performance
Many benchmarks adjusted retroactively to reflect the ACO's actual patient population treated during the performance period. Retroactive adjustments remove patients who were expected to be assigned to the ACO but received their care elsewhere. While this adjustment allows benchmarks to reflect where beneficiaries actually received their care, it also means ACOs do not know their true benchmark during the performance period. We support allowing voluntary prospective beneficiary assignment, which helps mitigate uncertainty and unexpected benchmark changes by relying on a predictable beneficiary population. We urge payers to allow an option for prospective beneficiary assignment, a process by which beneficiaries may attest that they consider a particular provider responsible for coordinating their overall care. An attesting beneficiary would be attributed to the ACO with whom that provider is affiliated, and that attestation would take precedence over other attribution methodologies. Voluntarily beneficiary assignment balances the important considerations of beneficiaries' freedom of provider choice with ACOs' interest in reducing patient turnover or "churn" and results in more defined and stable beneficiary populations and benchmarks, which are known up front. Further, prospective assignment allows ACOs to better target their efforts to manage and coordinate care for beneficiaries for whose care they will ultimately be held accountable.

Increase benchmarks as a result of high quality performance
Quality is an essential focus of population-based payment models, and payers should reward top performing providers with higher benchmarks for higher quality. This can be done by adjusting benchmarks or lowering minimum savings rates in relation to quality performance. ACOs that attain high quality should be rewarded, not merely avoid penalties. Further, to positively recognize exceptional quality performance, ACOs in the top quartile of quality performance should be eligible to earn higher shared savings, up to a 10 percentage point increase in shared savings on a sliding scale based on performance. In addition to evaluating quality performance compared to established measure thresholds, it is imperative to recognize – and reward – quality improvement relative to an ACO's previous performance. In fact, we recommend that quality performance and improvement be equally weighted, which would provide a strong incentive for ACOs to focus on improvement.

CONCLUSION

As noted in the white paper, establishing financial benchmarks is critically important and technically challenging. Without proper benchmarks, providers do not have adequate opportunity to succeed in population-based alternative payment models such as ACOs. We appreciate the opportunity to provide our feedback and look forward to continuing this important dialogue.

Sincerely,

Clifton Gaus CEO