Full Letter to Medicare on Innovation in Value Based Health Care

December 5, 2020

By Travis Broome

*This is the full 3,600 word CMS response letter. A summary of our thoughts is available on our blog.

November 20, 2017

Seema Verma, Administrator

Centers for Medicare & Medicaid Services

7500 Security Blvd

Baltimore, MD 21244


Re: RFI: Innovation Center New Direction

Dear Administrator Verma:

Aledade (aledade.com) partners with 272 primary care physician practices, FQHCs and RHCs in value-based health care. Organized into twenty accountable care organizations across 18 states these primary care physicians are accountable for over 240,000 Medicare beneficiaries. More than half of our primary care providers are in practices with fewer than ten clinicians. We are committed to outcome-based approaches to determine the value of health care. We are committed to using technology, data, practice transformation expertise and most important the relationship between a person and their primary care physician to improve the value of health care.

Future of Medicare and Medicaid Innovation in Value-Based Health Care


We appreciate the opportunity to respond to the Center for Medicare and Medicaid Innovation’s request for information (RFI) on Sept 20, 2017. We seek to share what physician practices have learned in the transition to value and our views on how to continue moving forward.

Physician-only ACOs face unique challenges in model participation while also outperforming other types of ACOs[1][2][3]. We view the RFI through the lens of models that are led by physicians.


Guiding Principles

  • Choice and Competition in the Market –Congress has taken initial steps to reduce regulatory incentives encouraging the merger of hospitals and physician practices, but more needs to be done. New models should further eliminate payments for physician practices to merge with hospital systems such as facility fees creating higher payment for the same services and the 340B program making drug pricing uncompetitive in private practice.


Competition is also key to success in value-based health care. CMS should prohibit anticompetitive behaviors such as data blocking and anti-tiering provisions that prevent the creation of financial incentives for using high-value health care providers.


  • Provider Choice and Incentives – A model that provides a business case for improving care will attract voluntary enrollment by physician practices. These models should, over time, put physician practices at financial risk, but that risk must be proportional to the finances of independent physician practice and not so large as to favor consolidation of practices. Models should move over time to a financial and evaluation structure focused on analysis of their local market. Physician practices should be able to accelerate the move to a difference-in-difference approach by taking on risk.
  • Patient-Centered Care – A strong primary care physician-patient relationship is the strongest tool available to create more value in health care. This proposition is strongly supported in the health services research literature and in the results of the MSSP.
  • Benefit Design and Price Transparency – Price transparency to model participants and to the consumer of health care creates competition by informing the choices of both beneficiaries and referring physicians. Benefit design should incentivize the building of the primary care physician-patient relationship and other cost-saving choices.
  • Transparent Model Design and Evaluation – Transparent design serves as a key component of voluntary provider choice. Uncertainty creates reluctance, but transparency promotes understanding and increases commitment. For evaluation, we favor difference in difference models of evaluation.
  • Small Scale Testing – All models should be designed with success and scalability in mind. We believe that Accountable Care Organizations can serve as unique test beds for innovation models due to ACO incentive to reduce total cost of care.


Focus Areas

  • Increased participation in AAPM – Three primary levers for physician practices:
    • Refine existing benchmarking methodology to better relate to local markets and reflect the health of the underlying population with accurate risk scoring
    • Calibrate risk to the finances of physician practices while still offering within-model reward for taking on risk for most models including all MSSP tracks
    • Create a new, more flexible, and sustainable version of Next Generation ACO for full risk taking with traditional Medicare
  • Consumer-Directed Care & Market Based Innovation Models – Allow Medicare beneficiaries to identify to CMS their primary care physician and create models that allow for beneficiaries to share in savings if their PCP participates in a savings model
  • Medicare Advantage Innovation Model – Facilitate innovative MA plans by allowing new physician-run plans to use existing Medicare infrastructure (e.g., MAC claims processing systems) so physicians can focus on population health management.
  • Physician Specialty Models – Focus on models that work within the framework of a total cost of care model as well as stand-alone models.
  • Prescription Drug Models – Test inclusion of Part D drugs in total cost of care models.
  • State-Based and Local Innovation – In addition to focusing on new models, CMMI can serve as a resource for physicians to learn about state and local innovation.
  • Mental and Behavioral Health Models – Use total cost of care models to test greater investment in mental and behavioral health.
  • Program Integrity – Allow physician practices, ACOs and other model participants to collaborate with CMS on surveillance and increasing transparency.

Increasing Participation in AAPM

Advanced Alternative Payment Models should be judged based on their ability to attract participants. The primary outcome measure for an AAPM should be how much value it created. The value is a combination of the percentage of savings or other outcome improvement measure times the number of people the model effected. There is no better way to encourage participation than a well-designed and well-understood model. This model relies on provider choice and transparency in development and is designed to attract willing physicians and other participants.

For physicians the crux of an advanced versus regular APM revolves around the phrase “more than nominal financial risk.” One of the first questions we get from our physicians is will the ACO qualify as an AAPM? Whether that risk is something that a given physician should undertakes revolves around how well the model benchmark separates risk due to the effective delivery of health care services and population health services versus risk that is due to uncontrollable circumstances or insurance risk. Aledade now partners with over 1000 primary care physicians who believe in population health and their role in it. However, they do not feel responsibility for events they can neither control nor influence. We have seen physicians and their staffs make great efforts to get the most non-compliant person into the office and out of the emergency department, on their medications and working towards their own health. However, no primary care efforts will influence whether that person develops unavoidable cancer. Nor will any primary care initiative account for regional differences in cost structures that have developed over decades. If you are financially responsible for whether a patient develops an unavoidable cancer you are an insurance company and that is a business that most physicians do not want to be in. All models should use risk scoring methodologies that accurately set targets, particularly advanced alternative payment models where the participant is taking risk. Models should acknowledge that changes in health will vary between model participants. This means that for a given model participant risk adjustment should be able to raise or lower the cost target. At the same time, we recognize that CMS has a vested interest in not rewarding model participants for changing their risk score more than the underlying health of the beneficiary population changes. At the program level this could take the form of either a cap in year over year change or a program wide adjustment factor. Regardless of the method CMS chooses to protect the program, an individual model participant’s cost target should track changes in their risk score (i.e. if the population is sicker the target higher, if the population is healthier the target lower) even if the magnitudes of the change are not 1:1 in order to protect the program at large.

The other way to increase the accuracy of model benchmarks is to relate them to local health care markets. Finally, since insurance risk cannot be completely eliminated the risk to which physician practices are exposed to must be more than nominal, but never ruinous. The goal of downside risk is to motivate the model participants and give the payer assurance that the ACO’s interests are aligned with the payers or in the case of Medicare society’s interests. Rather than setting downside risk in its current mostly symmetrical fashion just because it feels fair, CMS and other payers should set downside risk to accomplish the goal of motivation. Models like Track 1+, that relate risk to the finances of the participants instead of the model benchmarks will greatly encourage AAPM participation, particularly if those models offer rewards in the model for taking on risk, which is not currently the case with Track 1+.

Finally, we recommend that CMS dramatically reduce the timeline for when a physician is rewarded with the 5% bonus created by MACRA. Under the current timeline, a physician who makes a decision to join an AAPM in July of 2018 (most AAPMs have July deadlines) for 2019 participation will not see their 5% bonus until May 2021, nearly three years later. The physician could receive payments or pay losses based on their performance in the AAPM in September 2020 or nine months earlier than the bonus they get just for participation. This timeline is serving as a drag on AAPM participation. For AAPMs that require full year participation CMS could assume participation for 2019 as early as the finalization of the model participant list in December 2018 and then retroactively look at 2018 claims to determine whether the physician is a qualifying professional. This would allow CMS to move the MACRA AAPM bonus payment forward two years to spring 2019 and serve as a much better incentive for physicians to join an AAPM this summer. Principally, CMS should always be looking for ways to shorten the time frame from when physicians take action to when the outcomes of those actions are rewarded.

In conclusion, to encourage greater AAPM participation for 2019 we recommend:

  • Design new models and refine all existing models to focus on value creation that is within control of physicians and patients through local market benchmarking and accurate risk scoring as no incentive payment will ever overcome significant transfer of insurance risk from payer to provider
  • Require risk, but make the risk motivational and rewarding, not ruinous
  • Dramatically shorten the MACRA AAPM bonus timeline and continually seek to shorten the timeline between action and outcome
  • Create a new, more flexible, and sustainable version of Next Generation ACO for full risk taking with traditional Medicare

Consumer-Directed Care & Market Based Innovation Models

We believe one of the strongest tools to create value in health care is the relationship between primary care physician and patient. We suggest centering consumer-directed care around that relationship. CMS should create a mechanism for beneficiaries to select their primary care physician and encourage, but not require, them to do so. If a beneficiary selects a PCP who is an alternative payment model and that PCP succeeds in the model, the beneficiary should share in that success.

We suggest that this be in the form of reduced deductible and/or Part B premiums in the next year. We believe that by the patient should get the benefit automatically and that the ensuing discussion between patient and PCP on the success of the model will align incentives and create loss aversion on behalf of both parties that will benefit all involved. We are missing out on an opportunity when we only align incentives between health care providers and health care insurers. By creating an opportunity for Medicare beneficiaries to realize savings as well, we engage all parties in the effort.

In addition and not necessarily in conjunction, we recommend that CMS streamline the waiver process and expand it to allow model participants to offer reduced cost sharing to Medicare beneficiaries and to allow model participants to invest not just in health care related services such as home monitoring equipment, but also social services such as accessibility to the home and nutritional needs.

Medicare Advantage Innovation Model

Medicare Advantage is an excellent opportunity to align incentives, but it suffers from a lack of competition[4]. Since 1997, Medicare Advantage has tested the premise that the private sector can compete with Medicare in providing health care to seniors. Through its many iterations and refinements, two aspects of the program have never changed: First, to compete with Medicare, private companies must take over claims processing from the Medicare administrative contractors (MACs). Second, those companies must also create their own provider contracts. Twenty years since the program began, health care plan competition consists of much more than efficient claims processing and provider contract negotiations. Yet these capabilities are still two of the main barriers to entry into Medicare Advantage, blocking efficient and innovative providers from participating in the program. We propose removing these barriers by opening up Medicare Advantage to health care providers without a dependency on legacy plan capabilities such as claims processing and network contracting. We believe that the two greatest drivers of health care value are increasing and maintaining competition and aligning incentives of physicians and other health care providers with Medicare and with Medicare beneficiaries. Our proposal will do both by building a network on top of Medicare participation—not instead of it—and leaving claims processing in the highly experienced, efficient hands of the MACs. Making this change will shift the conversation about provider networks from price concessions and market power to creation of truly patient-centric, quality-based networks led by primary care. This will result in better care for patients, while allowing traditional Medicare to realize deeper savings through competition and aligned incentives. By removing traditional plan operations as a barrier to entry for Medicare Advantage, Medicare can create a path for successful provider groups to move into Medicare Advantage. This increase in competition will benefit both Medicare beneficiaries and health care providers. We discuss this proposal in greater details in Health Affairs[5]

CMS should also review current Medicare Advantage regulations to ensure they encourage participation in value based contracting. For example, the Performance Based Incentive Payment regulations have not been updated since the emergence of accountable care and other models.

Physician Specialty Models

CMS should explore models for specialists that integrate with total cost of care models and models that stand alone. As CMS works on physician specialty models, we recommend that it always consider the effects on other models. We further recommend that in cases where models do overlap that the model with the most risk would receive precedence in assign the financial outcomes of the model. In all cases, overlap should never result in a situation where costs are assigned to one model due to another model that are higher than historical costs for the model participant. For example, if a joint replacement model overlaps with a total cost of care model the joint replacement could be assigned the price of $20,000. So the total cost of care model gets assigned $20,000 no matter what the actual cost is. However, if the participants of the total cost of care model historically have an average cost of just $18,000 for a joint replacement, then the overlap is creating an artificial $2,000 loss. This situation must be avoided as it creates animosity rather than collaboration.


Mental and Behavioral Health Models

Total cost of care models are unique opportunities to test further investment in mental and behavioral health. When health care providers are responsible for total cost of care it removes incentives to cost shift and creates an incentive to maximize the volume derived from additional investment in mental and behavioral health.

Our experience with the Comprehensive Primary Care Plus (CPC+) program, specifically its Track 2, shows the difficulty in enabling behavioral health care in the primary care setting. Primary care physicians view this work as distinct and specialized requiring dedicated staff and specific expertise. To fund these activities, behavioral health is not served well by wrapping its funding into other funding streams. It is better for it to be clear to physicians what resources are dedicated to behavioral health. We recommend that CMS clarify CPC+ Track 2 behavioral requirements and consider a new model for behavioral health within total cost of care models.

For CPC+ Track 2, we recommend CMS issue a white paper or other education that ties a specific amount of the increase between CPC+ Track 1 and Track 2 to the behavioral health integration requirements. This would give practices much needed information to inform investment levels in behavioral health. This need not be a requirement, but simply filling an existing knowledge gap that we believe is holding back investment and participation in behavioral health integration with primary care.

For a new model of behavioral health, we recommend that CMS launch a behavioral health model that will provide new payment models within total cost of care organizations like ACOs and their participants. Structured around a predictable payment schedule, this investment can test behavioral health specific resources within the context of an organization that is responsible for the total cost of care. CMS could also test the effects of varying payment levels of existing behavioral health payments within a total cost of care organization. By allowing the organization to decide on how best to deploy the resources, either at the practice level or the organizational level, CMS creates flexibility for organizations of different make-ups to apply behavioral health integration as it best relates to their local health care market.

Increasing Value in Health Care through Innovation

We believe there are three main drivers of increased value in health care: competition, aligned incentives and professionalism. Competition increases value; however, it must be encouraged and even protected. Unlike professionalism and aligned incentives, competition does put downward pressure on health care provider’s margins creating an incentive for health care providers to find ways to reduce competition. CMS must always be aware of the effects that new models will have on competition. These effects are not always intuitive. It is possible for a model to both encourage further consolidation among large health care providers and also provide an avenue for independence for smaller health care providers. For example, the accountable care model is an opportunity for independent practices to take advantage of today’s advances in technology and data to help their patients navigate the whole health care system without needing to vertically integrate. It can also be a catalyst for a large integrated delivery system to complete their vertical integration efforts. To understand these sometimes competing effects, markets must be critically evaluated for competition and the make-up of model participants should be understood and categorized by type of provider physician practices, hospitals, other facilities, and integrated delivery networks. This does not always present itself as classical consolidation. For example, non-compete clauses in hospital employment contracts (eg that bar medical practice within 50 miles for 5 years) could be an anti-competitive abuse of market power that would not show itself in a traditional market analysis. Finally, models must also consider that health care providers do not just compete among each other they also compete with insurance companies for share of health care dollars. Models should consider the effects on competition with health insurers as well as competition among health care providers.

Aligned incentives is the most recent addition to value driven health care. It is a challenging task. In health care rarely is the need for a service known in advance by the consumer of that service. To address this and other uncertainties, consumers transfer risk to health insurers thereby created a three party transaction between the health care provider, the health insurer and the consumer of health care services. Further complication exists in the third-party administrator situation where the health insurer (the party that bears the uncertainty) and the party that contracts with health care providers are different. Finally, the quality of the health care service provided is hard to measure and the definition quality itself is subjective. These complications can create disparate incentives where a physician may be incentivized to perform more tests, while a health insurer wants fewer tests, a third party administrator gets a percentage of the tests and the patient has no idea how many tests they really need. Our challenge is to align incentives in such a way to maximize the value of a dollar spent on health care. In order to do so we should create models that reward health care providers for increasing the value of the health care dollar and ensure that health care consumers also receive that value through lower premiums over time or even through direct payments to consumers.

Medicine has always been blessed with a high level of professionalism. That professionalism must be respected as we seek to increase value through competition and aligned incentives less we lose the value it brings to health care. Physicians, and indeed most health care professionals, feel their autonomy is under threat from nearly all quarters. Poorly designed quality measures that either do not help a physician provide better quality or impose unjustified burdens for data collection distract from the physician/patient relationship. Pressure on the margins for physician services combined with ever-increasing administrative burden to document physician services assault private practice from both the revenue and cost side decreasing competition as physicians seek negotiating power to increase revenue and scale to spread out the administrative burden. As we align incentives and increase competition, we should do everything we can to ensure we do not lose value that is created by professionalism.


We appreciate the opportunity to comment on the future direction of CMMI. We believe there is incredible opportunity for CMS to continue to lead the movement towards value based payment in health care.

As CMS conceives of and evaluates new models we summarize our thoughts as:

  • Models should be attractive enough to independent physicians that they choose to participate and to take risk
  • Competition creates value as well and CMMI should consider the competitive effects of all models
  • Professionalism of health care providers creates value as well and should be preserved and respected

We look forward to continuing to work with CMS to increase the value of the health care dollar. Please contact me or Travis Broome ([email protected]) if you have any questions about our submission and/or we can be helpful to you and your staff as you explore new directions for CMMI.

[1] https://www.nejm.org/doi/full/10.1056/NEJMsa1600142#t=article

[2] https://www.nejm.org/doi/full/10.1056/NEJMp1709197?query=TOC

[3] https://www.ajmc.com/contributor/travis-broome/2017/10/cms-releases-medicare-shared-savings-program-2016-results

[4] https://www.commonwealthfund.org/~/media/files/publications/issue-brief/2015/aug/1832_biles_competition_medicare_private_plans_ib_v2.pdf

[5] https://www.healthaffairs.org/do/10.1377/hblog20170706.060925/full/