Marilyn Tavenner
Acting Administrator
Centers for Medicare & Medicaid Services
Department of Health and Human Services
Attn: CMS 3239-P
P.O. Box 8010
Baltimore, MD 21244-8010

Re: CMMI Bundled Payment Pilot

Dear Ms. Tavenner:

The Healthcare Financial Management Association (HFMA) would like to take this opportunity to comment on the Center for Medicare and Medicaid Innovation (CMMI) Bundled Payment for Care Improvement initiative (BPCI). HFMA appreciates CMMI's efforts to pilot a variety of bundled payment approaches for the Medicare population.

HFMA is a professional organization of more than 39,000 individuals involved in various aspects of healthcare financial management. HFMA has formed a primary advisory committee in the areas of health payment and delivery system reform. In addition to this advisory committee, experts within HFMA's local chapters across the United States provide additional insights and analysis in order to address the unique characteristics of health service organizations.

In 2008, HFMA convened a group of healthcare stakeholders representing payers, providers, employers, and patients to define the key principles that a reformed payment system must achieve. The group came to consensus around five key principles which are discussed briefly below:(1)

  • Quality: Payments should encourage and reward high-quality care and discourage medical errors and ineffective care. Wherever possible, payments should reward positive outcomes, rather than adherence to processes. In the absence of outcome measures, payment systems should reward the use of accepted practice and evidence-based processes and protocols that meet or exceed standards of quality and safety to promote optimal outcomes. Payers should not be responsible for payment to cover costs directly related to serious preventable medical errors.
  • Alignment: Payments should align incentives among all stakeholders to maximize the efficiency and coordination of health services based on accepted practice and evidence-based delivery models and protocols. Payment systems should stimulate and reward healthful behavioral choices and selection of value-based services by consumers related to prevention, primary care, acute care, and chronic disease management. Care decisions should be made through a shared decision-making process in which patients' values and preferences are identified and respected.
  • Fairness/Sustainability: Payment systems should balance the needs and concerns of all stakeholders. Payments should recognize appropriate total costs for the efficient delivery of healthcare services that are necessary and consistent with evidence-based care, high-quality/low cost provider benchmarks, and the advancement of medical science. Payment systems should accommodate payers' and purchasers' needs to allocate funds in a predictable, manageable fashion. In addition, consumers should have financial incentive to select high-quality, efficient care without being discouraged from seeking necessary and appropriate services. Finally, the payment system should be sustainable, providing a stable funding stream in the face of competing claims on public and private capital.
  • Simplification: Payment processes should be simplified, standard, and transparent. All parties should use payment methodologies, standardized at the national level, to reduce complexity. The payment methodologies should be transparent to those affected by them, and comply with privacy, security, and antitrust laws and regulations.
  • Societal Benefit: The resources needed to support broad societal benefits (i.e. medical education and research, indigent care) should be paid for explicitly. Similarly, payment systems should reward innovators who develop technologies, services, processes, and procedures that enhance safe, high-quality, and efficient care.

HFMA has long supported the concept of "value-based payment" and we believe bundled payment arrangements represent a significant opportunity to overcome the barriers embedded currently in the volume driven fee-for-service payment system and achieve the five principles outlined above. In an effort to help our members prepare for bundled payment, HFMA has engaged in extensive dialogues with participants in both the CMS ACE demonstration and the PROMETHEUS Payment Project (see resulting whitepapers -- Appendix I and II.)

Unfortunately, the programs outlined in CMMI's request for application (RFA) do not achieve these goals and are deeply flawed. As currently conceived, the CMMI BPCI initiative:

  • Misses an opportunity to align incentives for quality and cost-effectiveness among all stakeholders by ignoring the role beneficiaries play in their care.
  • Jeopardizes fairness and sustainability by providing insufficient financial support for participating providers, given the risks involved in managing a bundle, and fails to provide sufficient reward for innovative delivery systems.
  • Increases complexity and the administrative burden on providers. Fails to adequately describe the legal waivers provided to create the alignment necessary for providers to coordinate care and improve outcomes.

Unfortunately, these flaws are not unique to the BPCI. As we have documented in previous comment letters and summarized in Appendix III, many of the value-based programs launched by CMS since passage of the ACA suffer similar faults.

Please find our detailed comments related to the specific components of the BPCI that concern our members below. We believe strongly in the need for payment reform in order to realign incentives and would welcome any opportunity to work with CMMI and CMS to reshape the BPCI and other value-based initiatives into ones that are beneficial for patients, providers, and the Medicare Program.

The Role of Medicare Beneficiaries
Given that BPCI participants will be selected to participate in the program based on their perceived ability to provide high-quality care in a cost-efficient manner, HFMA believes that all Medicare beneficiaries should be encouraged to receive their care, for selected conditions, from providers in the beneficiaries' region who have contracted with CMMI for those bundles. Otherwise an opportunity to provide safer, higher quality care to Medicare beneficiaries will be lost.

Further, beneficiaries who are attributed to a network of providers for conditions bundled under models two and three should receive all of their care from providers in the "contracted" network. This not only has important implications for the quality of care delivered to Medicare beneficiaries but also for the financial sustainability of the contracted networks that are taking risk in an effort to reduce overall healthcare spending. Unfortunately, research shows that nationally only 48 percent of Medicare beneficiaries will always stay with a group of providers that is accountable for their care.(2)

HFMA believes the lack of useful quality information and positive incentives for beneficiaries to use BPCI providers and stay "in-network" for all of their care poses significant financial risk to participating organizations. First, it is unreasonable to ask a network of providers to assume downside risk for a given condition (particularly for models two and three) if they cannot ensure care will be delivered based on evidence-informed best practice protocols by providers who are working under financial incentives that are aligned to support coordinated care.

Second, given the significant financial risks that BPCI participants are assuming under the CMMI program, they should be rewarded with higher patient volume or in other ways to mitigate the downside risk. Unfortunately, as we have discussed in an earlier comment letter, both the CMS shared savings program and the CMMI Pioneer ACO model suffer from similar design flaws (see Appendix IV).

We believe that CMMI should take the following steps to help beneficiaries receive the highest quality care possible while helping participants manage risk and increase their volume:

  1. Provide beneficiaries, on an annual basis, with a listing of BPCI networks in their area that includes all associated providers by provider type and the conditions for which each network has contracted. The listing should include a comparison of easy to understand quality metrics between the BPCI contracted provider and other facilities in the community.
  2. Allow participating providers to market themselves as CMMI/CMS "Quality Centers" for the conditions they've contracted with CMMI.
  3. At a minimum, calculate beneficiary cost sharing based on the price of services included in the "bundled bid." However, an optimal solution would use the authority granted by the Section 3021 of the ACA to waive beneficiary cost sharing requirements for those beneficiaries who receive care for bundled services from BPCI participants.

Taking these steps will help CMMI achieve a number of its stated goals and avoid potential complications with the program. Some of the desired outcomes are noted below.

  • Patients will be encouraged to seek out care from providers that CMMI has determined provide high-quality, coordinated care. It's only right that patients share in the savings generated by active participation in their own care. This will also alleviate potential concerns that the Medicare program and BPCI participants stand to benefit from rationing care provided to beneficiaries.
  • The volume of patients in the program will increase as beneficiaries seek out BPCI participants to receive the financial benefit when they require care for a given procedure or condition. This will directly benefit CMS as it will accelerate the "bending of the cost curve." Additionally, it will reward providers who participate in the program, ensuring their financial stability, and thereby encouraging more organizations to participate.
  • Organizations will be able to better manage care and understand utilization patterns in real time if beneficiaries consistently stay within the bundled network for all required services. Reports run from the clinical and financial decision support systems will provide an immediately available, more accurate snapshot of beneficiary needs, allowing the organization to more effectively reengineer care processes. This is particularly true for models two and three.

Degree of Risk Assumed by BPCI Participants vs. Potential Reward
HFMA is concerned that the design of the BPCI will transfer insurance risk to participants. We believe this was clearly not CMMI's intention when they designed the BPCI and is certainly not acceptable from the participant/provider perspective. We strongly encourage CMMI to take the following steps to address the potential for inappropriate shifting of insurance risk to potential BPCI participants.

Beneficiary Engagement: We believe that, assuming BPCI participating providers provide all of the care for a given bundle, it is appropriate for them to assume performance risk. However, given the lack of incentives (as discussed above) for beneficiaries to stay "in-network" for a given bundled service, it will be difficult to ensure that beneficiaries receive optimal care. Care received outside of the "contracted network" (for which models two and three are at greatest risk) will be delivered by providers who are not operating under the same financial incentives and have not used evidence-based medicine to ensure that high-quality care is delivered in a cost-efficient manner. Unless this situation is rectified, we believe it will lead to sub-optimal outcomes for beneficiaries at a higher cost to both the beneficiary and the Program. To address this issue we strongly encourage CMMI to implement the recommendations discussed in the previous section.

Risk Adjustment Mechanism: HFMA commends CMMI for allowing BPCI applicants to propose their own risk adjustment mechanism. However, we are concerned that CMMI will not accept models that take beneficiary economic status into account. Our concern is based on the position CMS articulated in the 2012 IPPS final rule related to the Hospital Readmissions Reduction Program (RRP).

HFMA believes that CMS, in the instance of the RRP, is ignoring the preponderance of evidence that inclusion of variables such as Medicaid as Secondary Payer and SSI status significantly improves the accuracy of risk adjustment. Given the predictive power variables such as the presence of Supplemental Social Security Income (SSI) have related to readmissions and other outcomes(3) that are not included in the adjustment mechanism for the readmissions program, HFMA recommends that CMMI, similar to the CMS Shared Savings Final Rule, use the CMS-HCC or similar model to allow for the inclusion of SSI and other similar socioeconomic indicators (e.g., presence of Medicaid as a secondary payer) to improve risk adjustment in the BPCI. This is particularly crucial for models two and three, which bundle payments over an extended episode.

Further, recent analysis(4) has shown that safety net hospitals are more likely to have higher readmission rates. We continue to believe that allowing a refined risk adjustment mechanism is necessary to encourage safety net hospitals to participate in initiatives like the BPCI that will improve the quality of care provided to beneficiaries and ultimately lead to better outcomes at a lower cost to both the beneficiary and the program.(5)

In the future, as more providers adopt EMRs, HFMA also recommends that CMS and CMMI explore ways to use the embedded data to refine the risk adjustment mechanism and improve the accuracy of readmissions models. Examples of variables for consideration in the future should include housing discontinuities as measured by address changes, census tract, history of drug use, and marital status.

Maximum or "Capped" Losses: HFMA is deeply concerned the BPCI doesn't have explicitly stated caps on losses. This concern is exacerbated considering that participating providers have already offered significant discounts off the "Medicare Fee Schedules" for all services included in the bundle and are not receiving any type of "preferred network status" through marketing or beneficiary financial incentives as discussed above. Like all other CMS value-based purchasing arrangements, HFMA believes the BPCI needs clearly articulated loss caps calculated at the MS-DRG level. We believe that much like the shared savings program, loss caps should be phased in at 5 percent in the first year and increase to a maximum of 10 percent of the accepted episode "price," trended forward to the current year.

Further, we are concerned about the statistical reliability of the data sets available to providers in smaller markets (i.e., Medicare population of 5,000 or less). As CMS knows from experience, bids based on analysis of smaller data sets will be subject to higher volatility in actual results. In order to encourage providers in smaller markets to participate and shield them from unintended consequences we believe that:

  • Loss caps for providers in smaller markets should start at three percent of the accepted episode price trended forward to the current year and be phased in to a maximum of seven percent.
  • Small market providers should have an opportunity to renegotiate their bid based on actual results if they substantially deviate from the CMMI accepted episode price.

Outlier Cases/Unbundled Services: In addition to capping losses, CMMI offers a mechanism to account for "outlier cases" and "unbundled services" by allowing applicants to identify and exclude conditions and services from the proposed bundle. This approach suffers from two significant flaws:

  • It is impractical (especially given the limited, two year data set provided to applicants) for a provider to list every low-probability but potentially costly clinical scenario for exclusion from a proposed bundle.
  • CMMI will likely standardize the bundles for each MS-DRG based on commonalities found in the application as opposed to administering multiple bundles for the same model and MS-DRG, thereby exposing some providers to increased outlier risks.

To mitigate the problems discussed above, in addition to capping losses and allowing for stated exclusions from the bundle, HFMA recommends that CMMI exclude cases that fall at or above the 95th cost percentile for the participant's market from the reconciliation calculation.

Post-Period Monitoring
: HFMA understands the need for a post-episode monitoring period to ensure care isn't delayed and costs shifted. However, we believe these concerns are overblown. If providers were to purposefully delay necessary care, it would likely decrease quality across a variety of process and outcome metrics that we anticipate CMMI will monitor.

We believe that monitoring cost benchmarks is the wrong approach to ensuring that appropriate care is provided. The following is an extreme example to highlight a key point. Improvements in care delivery have the potential to decrease mortality rates for some participating providers. Beneficiaries who previously could have expired because of insufficiently coordinated care could incur increased healthcare costs in the post-episode period due to their frail condition. This desired outcome may not be accounted for in the risk-adjustment mechanism and the participant, who significantly improved beneficiary outcomes, could be financially penalized.

Beyond the undesirable example outlined above, there are several overall design elements that transfer insurance risk to providers:

  • The open-network nature of the program exposes providers to additional, unnecessary risk in the post-episode period if cost benchmarking is used to ensure appropriate care delivery. As discussed above, it is unrealistic to hold a group of contracted providers responsible for the efficiency of care delivery if they are not providing the care.
  • Exact dates of service are not available in the data files provided to applicants. Not only will it be difficult to construct relevant episodes and bid on them for models two and three, but it will also be impossible for a provider to fully understand its risk exposure in the post-episode monitoring period.
  • There is little transparency as to how CMS will calculate the "risk threshold" in the post-episode period. Applicants will need to know both the statistical probability of exceeding the threshold solely by reason of chance and the length of time an applicant will remain at risk for CMS's computation of whether there were post-episode costs.

To mitigate these concerns, we strongly encourage CMMI to work with participating providers to identify a set of quality and outcome metrics to monitor instead of using cost/service utilization as a proxy to determine if optimal care was delivered at the appropriate time. We further encourage CMMI to adopt an approach where providers are given an opportunity to address any performance-related issues prior to the imposition of a financial penalty.

Data Sharing: Within the BPCI request for application there is no discussion of if or how CMMI will share claims data with participants during the contract period. We believe that claims data is an important tool that providers need when reengineering care delivery to improve quality and efficiency - particularly for models two and three where there is substantial risk of care provided out of network.

HFMA strongly recommends that CMMI use a process similar to the CMS shared savings program to provide BPCI participants with claims level data on a monthly basis. At a minimum, the data elements should include: Procedure codes, Diagnosis codes, Dates of service, Provider/supplier ID, and Part D data. Further, as we discussed in our comment letter on the readmissions final rule, we believe that all hospitals should have access to beneficiaries' Part D data in real time. This will allow for accurate medication reconciliation and significantly reduce preventable readmissions.

Administrative Complexity
The backend administrative processes are not discussed in the request for application. However, HFMA is concerned that if they are not designed properly they could unnecessarily burden participating providers by increasing the cost of the program without improving outcomes or the patient experience of care. As an example, we hear from ACE Demonstration participants that they continue to experience challenges related to:  

  • Reconciling their internal data to the reports that receive from the MAC
  • Timely administration of the Notice of Admission process with the MAC (implications for Model Four)
  • Having claims processed in a timely manner due to unnecessary claims edits

As a result of these and other issues, ACE participants have at least one or more FTEs dedicated to resolving these issues. For a small volume of cases, this significantly increases the administrative expense of the program.

We strongly recommend CMMI involve ACE participants and other providers who have expressed an interest in participating in the BPCI in the design of the administrative process. We believe this collaboration will result in a program that is more efficient and less onerous for both the participants to manage and the Program to administer.

Communication with Medigap Payers and Medicaid Programs: One of the many challenges ACE participants continue to encounter is adjudicating "secondary" claims for beneficiary deductible and co-payments. While CMS conducted some payer education at the outset of the project, it was insufficient.

We strongly recommend CMMI conduct continuous payer educational activities. Additionally, we would encourage CMMI to identify punitive steps it could take for recalcitrant Medigap plans and Medicaid Programs. Alternatively (as mentioned above), CMMI could address this issue by using its Section 3021 authority to waive beneficiary cost-sharing requirements for those who receive care for bundled services at BPCI participating providers.

Unresolved Questions: HFMA would like to commend CMMI for the amount of detailed information disseminated to potential BPCI applications through both the website and webinars. However, there are a number of key questions that remain unresolved. It would be beneficial for our members if CMMI would provide additional clarity around the following areas:  

  • CMS/CMMI ACOs and Bundled Payments: How will participation in both the BPCI and Shared Savings or Pioneer programs impact cost and risk profiles and annual settlements?
  • Value-Based Purchasing: How will participation in the BPCI impact the value-based purchasing program's potential penalties and bonuses? Will bonus payments be included in the per episode reconciliation to the "bid benchmark?"
  • Medicare Bad Debt Payments: Will bad debts for bundled services continue to be reimbursed through the cost report? Will participation impact any of the documentation requirements?
  • Full Risk Amount: In the FAQs related to participants' ability to bear risk, it states that participants must be able to bear the "full risk amount." How does CMMI define that amount?
  • Dispute Resolution: What mechanism will CMMI use to resolve disputes, particularly related to financial reconciliation, with participants?
  • Program Termination: On what grounds are participants allowed to terminate their BPCI agreement with CMMI? What process will they have to use to do so? Will changes in providers who participate in the BPCI under a contract impact an organization's ability to continue participating?

Legal Waivers: The BPCI request for application states that "The Secretary will consider exercising this waiver authority (Section 3021 of the ACA) with respect to the fraud and abuse laws in Titles XI and XVIII as may be necessary to develop and implement the Bundled Payments for Care Improvement initiative. The Secretary may also consider waiving additional provisions under Title XVIII for this purpose."

HFMA strongly encourages the Secretary to use the waiver authority accorded her under Section 3021 of ACA to create safe harbors for BPCI participants from fraud and abuse laws in Titles XI and XVII. These should mirror the five waivers created for the shared savings program (pre-participation, BPCI participation, gain-sharing distribution, compliance with physician self-referral, and patient incentive waiver). Further we would also encourage the CMMI to work with other regulatory bodies such as the FTC/DOJ and the IRS to ensure that participation in the program will not expose participants to anti-trust or private inurnment issues.

Closing Thoughts
HFMA commends CMMI's efforts to create a payment system that encourages quality, aligns incentives for all stakeholders, is fair and sustainable, is simple to administer, and funds necessary societal benefits. However, the BPCI, like many of CMS's other efforts, falls short on each of these aims across multiple provisions, as discussed in each of the sections above. Without significant revision, we are concerned the BPCI will not attract as many participants from the provider community as are interested in exploring bundled payment methodologies. For the ones that do elect to participate, we have significant concerns about detrimental financial impacts, which will limit their ability to continue participating. Under both scenarios, a significant opportunity to improve quality and increase the cost-efficiency of patient care for beneficiaries will be missed.

HFMA looks forward to any opportunity to provide assistance or comments to support CMMI's efforts to create a bundled payment program. As an organization, we take pride in our long history of providing balanced, objective financial technical expertise to Congress, CMS, and advisory groups.

We are at your service to help CMMI and CMS gain a balanced perspective on this complex issue. If you have additional questions, you may reach me or Richard Gundling, Vice President of HFMA's Washington, DC, office, at (202) 296-2920. The Association and I look forward to working with you.


Richard L. Clarke, DHA, FHFMA
President and Chief Executive Officer
Healthcare Financial Management Association

Cc: Valinda Rutledge, Director of Patient Care Models Group, Center for Medicare and Medicaid Innovation  


  1. Healthcare Financial Management Association, Healthcare Payment Reform - From Principles to Action
  2. Top Health Industry Issues of 2011: Health Reform Prompts Industry Players to Undergo Makeovers, PricewaterhouseCoopers, December 2010
  3. S. Jencks et al., "Rehospitalizations"
  5. Kansagara, D., Englander, H., Salanitro, A., Kagen, D., Theobald, C., Freeman, M., Kripalani, S. "Risk Prediction Models for Hospital Readmission: A Systematic Review." JAMA, October 19, 2011

About HFMA

The Healthcare Financial Management Association (HFMA) provides the resources healthcare organizations need to achieve sound fiscal health in order to provide excellent patient care. With more than 39,000 members, HFMA is the nation's leading membership organization of healthcare finance executives and leaders. HFMA helps its members achieve results by providing education, analysis, and guidance, and creating practical tools and solutions that optimize financial management. The organization is a respected and innovative thought leader on top trends and challenges facing the healthcare finance industry. From addressing capital access to improved patient care to technology advancement, HFMA is an indispensable resource on healthcare finance issues.

Appendix I: Pursuing Bundled Payments Lessons from the ACE Demonstration
Appendix II: Transitioning to Value PROMETHEUS Payment Pilot Lessons
Appendix III: Comparison of HFMA Payment Principles to CMI-CMMI Value Based Reimbursement Programs
Appendix IV: Proposed Rule on the Medicare Shared Savings Program: ACOs and Medicare Program: Waiver Designs in Connection With the Medicare Shared Savings Program and the Innovation Center; Proposed Rule and Notice

Publication Date: Monday, May 21, 2012