Monday, March 02, 2020

How CMS Can Create More Successful APMs More Quickly

In 2010, Congress created the Center for Medicare and Medicaid Innovation (CMMI) and provided it with significant funding and regulatory flexibility to develop and test Alternative Payment Models (APMs). In MACRA, Congress created incentives for physicians to participate in Medicare APMs and specifically encouraged the creation of more physician-focused APMs. 

Unfortunately, a decade after creation of CMMI and five years after passage of MACRA:

  • CMS has not created APMs for most Medicare beneficiaries or physicians. The majority of Medicare beneficiaries and the majority of physicians are unable to participate in any Medicare APMs. There is no Primary Care-Focused APM available for most Medicare beneficiaries and their primary care providers, there are no Condition-Based APMs designed to support better ambulatory care by specialists for patients with chronic conditions, and there are no APMs designed for most patients receiving outpatient procedures.  (A summary of current CMS APMs is available here.)
  • CMS APMs have failed to produce significant savings or improvements in quality. To date, most CMMI APMs have actually increased Medicare spending and have not resulted in significant improvements in the quality of care. ACOs in both the Medicare Shared Savings Program and the Next Generation ACO program have failed to achieve significant savings.  (A summary of the impacts of CMS APMs is available here.)

The failure to implement successful APMs means that each year, millions of Medicare beneficiaries are being denied the opportunity to receive higher-quality care and the Medicare program is spending billions of dollars more than is necessary.

The Problems With the CMS Approach to Creating APMs

There are four basic reasons why there are so few successful Alternative Payment Models in Medicare: 

  • CMMI has used only a small portion of its funding from Congress for development and implementation of Alternative Payment Models. Over the past decade, less than half of CMMI’s total spending, and less than 15% of the funds Congress appropriated for CMMI projects, was used to develop and implement APMs in which physicians could participate under MACRA. Moreover, nearly one-third of the funds spent to develop APMs were used to create variations on ACOs and the Medicare Shared Savings Program, rather than to create completely different types of APMs in which specialists and small and rural physician practices could participate.  (A summary of how CMMI has spent its funds is available here.)
  • CMMI uses a slow, expensive process to design and implement APMs.  The steps that CMMI follows in choosing, designing, implementing, and evaluating APMs require 7-9 years or more to complete, and CMMI has spent $75 million or more solely on model design and evaluation contracts for each one of the APMs it has developed. The cost and time involved in this process makes CMS less likely to test multiple APMs, causes unsuccessful APMs to continue operating for too long, and prevents successful APMs from being created and expanded more quickly.
  • The APMs that CMS has developed fail to solve the problems with current payment systems. Most of the APMs created by CMS are merely pay-for-performance programs: no changes are made in current fee-for-service payments, and providers receive bonuses or penalties based on whether spending is less than CMS projections. These types of APMs have not given providers the resources or flexibility they need to deliver high-value services to patients that could reduce spending and improve quality. Moreover, these types of APMs can penalize providers for things they cannot control and reward providers for failing to deliver services patients need. There is no evidence that simply increasing the amount of downside risk in APMs will produce greater savings.
  • CMS has refused to implement additional or different APMs. More than 30 proposals for Alternative Payment Models have been developed by physicians, medical specialty societies, health systems, and other individuals and organizations and submitted to the Physician-Focused Payment Model Technical Advisory Committee (PTAC) created by Congress. PTAC has recommended that 16 of these proposals be implemented or tested, but CMS has not implemented or tested any of them and has no plans to do so.  (A list of the APMs recommended by PTAC is available here, and an analysis of the reasons CMS has given for not implementing them is available here.)

How to Create More Successful APMs in Medicare

Clearly, a different approach to creating APMs in Medicare is urgently needed. There are three things that CMS can and should do to accelerate the implementation of more APMs that will achieve much greater savings for the Medicare program and improve the quality of care for many more Medicare beneficiaries:

#1 Design and Implement Patient-Centered Alternative Payment Models. Rather than continuing to add more “incentives” and “risk” on top of current fee-for-service payment systems, CMS needs to take a patient-centered approach to designing APMs.  A four-step process should be used for designing a Patient-Centered Alternative Payment Model:

Step 1: Identify one or more specific opportunities for reducing spending and/or improving the quality of care for Medicare beneficiaries;

Step 2: Identify the changes in care delivery that will reduce spending or improve quality in those opportunity areas;

Step 3: Identify the barriers in the current payment system that prevent or impede implementing the improved approach to care delivery;

Step 4: Design the Alternative Payment Model so that it removes the barriers in the current payment system and assures the delivery of higher-value care.

A Patient-Centered Alternative Payment Model developed through this process should have four key components:

Component #1 removes the barriers in the current payment system that prevent providers from delivering higher-value care. 

Component #2 requires accountability from participating providers for reducing aspects of spending that they have the ability to control. 

Component #3 requires accountability from participating providers for maintaining or improving aspects of care quality and outcomes they can control. 

Component #4 defines the patients who are appropriate for the services supported by the APM and ensures they are willing to participate before services begin. 

There is no single Alternative Payment Model that will work for all types of patients and all types of healthcare providers, so multiple APMs will be needed to successfully reduce spending.  

#2  Use a “bottom-up” instead of a “top-down” approach to creating APMs.  Physicians and other healthcare providers are in the best position to identify specific opportunities to reduce spending and improve quality for patients and to know what changes are needed in current payment systems to support higher-value care. CMS should encourage a greater role for healthcare providers in the development of APMs through the following actions:

  • Send a clear signal that well-designed APMs developed by physicians and other stakeholders will be tested by CMS. One obvious way to do this would be to implement at least a subset of the 16 APMs that have been recommended by PTAC. 
  • Provide data and technical assistance to help physicians and other stakeholders develop good APMs. HHS should either permit PTAC to provide data and technical assistance to APM developers or create a separate mechanism for doing do. 
  • Work collaboratively with providers to refine the details of APMs and encourage participation. 

#3 Use a more efficient and effective approach for testing APMs. Similar to the approaches used to encourage innovations in other industries, CMS should select multiple APMs for “beta testing” in order to refine the APMs and determine if they are likely to work before inviting large numbers of providers to participate and committing large amounts of money to extensive evaluations. This would enable design and testing of an APM to be completed within a 4-5 year period rather than the 7-9 years required under the current approach, and it would provide a much higher return on the investment of the funding Congress has made available.

Two modifications are needed to this approach in order to support successful primary care payment reform:

  • Testing of additional primary care APMs should be designed so that every Medicare beneficiary in the country has an opportunity to participate in a patient-centered primary care APM that provides adequate, flexible payments to support high-quality services. 
  • Because the biggest benefits of improved primary care will occur beyond the time periods typically used for evaluation, and because the short-term savings for Medicare may not offset the higher payments needed to support good primary care, Congress will need to change the law so CMMI can continue primary care APMs that improve the quality of care even if they do not reduce short-term spending.

More details on the problems with current CMS Alternative Payment Models and the steps needed to create more successful APMs in Medicare more quickly are available in the CHQPR report How to Create More Successful Alternative Payment Models in Medicare.


 Comments (0)

Thursday, December 19, 2019

How Should We Pay for Cancer Care?

In November, the Center for Medicare and Medicaid Innovation (CMMI) released the initial details of “Oncology Care First,” its proposal for a new Alternative Payment Model (APM) for cancer care that would replace the current Oncology Care Model. Cancer patients and those who care about them should be worried, because if Oncology Care First is actually implemented, it could prevent cancer patients from receiving the treatments they need and cause community oncology practices to close.

Over the past three years, the CMMI Oncology Care Model (OCM) has helped many Medicare beneficiaries with cancer to receive better care by filling in a key gap in the Medicare fee-for-service system — the lack of payment for care management services. The Monthly Enhanced Oncology Services (MEOS) payments provided in OCM represent more than $1,000 per patient in additional revenue for the nearly 200 oncology practices across the country participating in OCM, and the physicians have used that revenue to provide greater support for their patients and to respond more quickly and effectively when patients experience side effects from their treatment so they don’t need to go to an emergency department or be hospitalized.

But there is a second part to the Oncology Care Model — the “Performance-Based Payment” — and it has a very problematic structure:

  • OCM Creates a Financial Incentive to Withhold Needed Treatment.  An oncology practice can receive a bonus through the Performance-Based Payment (PBP) for withholding the delivery of an expensive treatment that a patient needs.  The quality component of OCM does nothing to prevent an oncology practice from stinting on care.
  • OCM Penalizes Practices for Using Evidence-Based Care and Encourages Practices to Avoid Patients Who Need More Expensive Treatments.  The methodology CMS uses to set Target Prices fails to adjust for important clinical differences between patients, changes in evidence about effective treatments, and large increases in the prices of drugs, and CMS reduces all Target Prices by an arbitrary “discount.”  This means that if a practice treats patients based on the most current evidence, spending will likely exceed the Target Prices, which would penalize the oncology practice financially.
  • OCM Rewards Practices for Delays in Completing Treatments.  In OCM, CMS pays for services in six-month “episodes,” which means they will receive significantly higher payments if they stretch out patient treatments to last longer than six months.
  • OCM Encourages Oncology Practices to Avoid Patients Who Have Health Problems Unrelated to Cancer Treatment.  The Performance-Based Payment in OCM is determined based on total spending on all services that an oncology practice’s patient receives for all of their health issues, not just services related to their cancer.  An oncology practice will be less likely to receive a Performance-Based Payment, and more likely to have to pay a penalty, if it has a higher-than-average number of patients with other health problems.

Even though the OCM Performance-Based Payment creates these undesirable incentives, oncologists were able to safely ignore them in the initial years of the demonstration project because none of the oncology practices signed up for the “downside risk” tracks in OCM.

However, that is now changing. Oncology practices are being forced to either (a) join one of the downside risk tracks where they would have to pay penalties to CMS based on this flawed methodology, or (b) exit the OCM program altogether and lose the additional MEOS payments they have been receiving.  Either way, the loss of revenues will likely result in reduced services and poorer outcomes for patients. 

Rather than building on the success of OCM by filling in additional gaps in fee-for-service payments and fixing the problematic Performance-Based Payment methodology, “Oncology Care First” would go in the exact opposite direction. It keeps the same Performance-Based Payment Methodology, forces every practice to pay penalties to CMS if spending on their patients exceeds the unrealistic Target Prices set by CMS, eliminates the MEOS payments and replaces them with unspecified new “Enhanced Services Payments,” and eliminates the payments all practices receive for office visits and chemotherapy administration, replacing them with a complex new “Monthly Population Payment.”

As a result, Oncology Care First has the potential to make cancer care worse by penalizing physicians for using new cancer treatments and paying them bonuses if they withhold the treatments that cancer patients need. The magnitude of the potential penalties is so large that they could force oncology practices out of the program or out of business entirely.

Cancer patients and their physicians shouldn’t be forced to choose between the flawed fee-for service system and an even more flawed alternative payment model like Oncology Care First. Fortunately, there’s a better way. Instead of Oncology Care First, CMS should implement Patient-Centered Cancer Care Payment with the two key components of a good APM: .

  1. New payments specifically designed to support the kinds of high-value services cancer patients need; and
  2. Accountability for reducing avoidable spending, so that savings can be generated without harming patients.

The details of how to create Patient-Centered Cancer Care Payment are explained in A Better Way to Pay for Cancer Care: The Problems with CMS Oncology Payment Models and How to Create Patient-Centered Cancer Care Payment. This report also describes the problems with current fee-for-service payments that need to be fixed and provides more detail on the serious problems with the Oncology Care Model, Oncology Care First, and the CMS Radiation Oncology (RO) Model.

Although changes are clearly needed in current payment systems to achieve higher-value healthcare, the cure shouldn’t be worse than the disease. Forcing physicians to take financial risk for the total cost of care for cancer patients will harm the patients and accelerate the loss of high-quality physician practices. It’s time to embrace a different approach to value-based payment for cancer care– a patient-centered approach instead of an payer-centered approach focused solely on reducing spending. 


 Comments (0)

Monday, May 13, 2019

The Disappointing Details About the CMS “Primary Care First” Payment Model

The most important element of a truly “value-based” healthcare system is strong primary care.  The reason is simple – the lowest spending and the best outcomes occur when patients stay healthy, and primary care is the only component of the healthcare system that is specifically designed to help patients prevent health problems from occurring and to identify and treat new problems as early as possible.

Unfortunately, the nation’s primary care system is at risk of collapse.  There is a large and growing shortage of primary care physicians in the country; many primary care physicians are burning out, and most medical students don’t want to go into primary care.  Although there are multiple causes for this, a major reason is the failure of the current payment system to provide adequate resources to support high-quality primary care services.  The problems are particularly severe for small primary care practices, which deliver most of the care in rural areas of the country.

In April 2019, the U.S. Department of Health and Human Services announced the “CMS Primary Cares Initiative,” consisting of five new payment model options intended to “transform primary care to deliver better value for patients throughout the healthcare system.”  The full specifications of the new Primary Cares Initiative options have not yet been released, but based on the details revealed so far, it appears they may fall far short of what is needed to fully address the problems facing primary care and to successfully sustain a high-value primary care system. 

The Problems with Primary Care First

Among the five options in the Primary Cares Initiative, “Primary Care First” is the only option that a small primary care practice will be able to participate in.  The “Direct Contracting” options are only available to practices that have at least 5,000 Medicare patients, which is far more patients than solo and small primary care practices will have.  In fact, most of the counties in the United States don’t even have 5,000 Medicare beneficiaries living in them.

Unfortunately, there are nine major problems with Primary Care First that prevent it from creating the kind of payment system that small primary care practices need in order to deliver high-quality care to their patients and stay afloat financially:

1. Practices Would Still Receive a Significant Portion of Revenues Based on the Number of Face-to-Face Office Visits.  Primary care practices have called for elimination of a payment system that pays for office visits and little else, but instead of doing this, Primary Care First actually creates a brand-new $50 fee for each face-to-face office visit.  Although it also would provide a flexible $24 per-patient per-month payment, more than 40% of a typical practice’s payments would still be tied to face-to-face visits.  Under Primary Care First, a practice that is able to care for patients effectively with fewer office visits will lose revenue and could be unable to cover its costs.

2. Payments for Many Patients Would Not be More Flexible at All.  A practice would only receive the new, flexible monthly payment for patients who are “attributed” to the practice based on where they received care in the past, and new patients would not be attributed for up to two years.  Patients could “voluntarily align” with the practice, but this requires the patient to create an account on the CMS website and go through a multi-step process to designate the PCP as their primary clinician.  Even if the patient successfully completes this process, the practice will have to wait 3-6 months to begin receiving the new payments. 

3. Practices Would No Longer Receive a Higher Payment for a Patient Who Has Greater Needs.  In the current fee-for-service payment system, a primary care practice receives more payment when it provides care to a patient with greater needs.  But under Primary Care First, a primary care practice would receive the exact same monthly payment and the same office visit payment for a patient regardless of how sick or healthy the patient is. 

4. Most Practices Would Receive No More Revenue Than They Do Today, and Less Revenue Than Under Other CMS Primary Care Models.  While opinions differ about the best methodology for paying for primary care, there is widespread agreement that primary care practices need to be paid more than they are paid today.  But CMS has indicated that the combination of the PBPM payments and office visit fees in Primary Care First are intentionally designed to provide no more revenue than current fee-for-service payments.  Despite receiving no additional revenue, however, a Primary Care First practice would be expected to deliver expanded services, including 24/7 access to a care team member, care management services, and integrated behavioral healthcare services.

5. Most Practices Would Receive Little or No Reward Based on Their Performance.  The “Performance-Based Adjustment” in Primary Care First could increase the amount a practice is paid by as much as 50%.  However, it appears that the majority of practices would only be able to receive a 3.5% increase in their total payments.  At most 10% of participating practices would be able to receive a 50% increase in payment, while every practice could have its payments cut by 10% if it fails to achieve high scores on quality measures or if its patients are hospitalized frequently. 

6. Accountability Measures Are Not Patient-Centered.  Under Primary Care First, delivering high-quality care would merely enable the practice to avoid a reduction in payment; the practice would only qualify for an increase in payment if its patients were hospitalized less frequently than the patients in other practices.  Moreover, a hospital admission for injuries in a traffic accident, for planned surgery to treat cancer, or for complications of chemotherapy administered by an oncologist would be treated the same as an avoidable admission for a chronic disease exacerbation, putting the practice at risk for things it can’t control.  The heavy weight placed on hospitalization rates is presumably designed as an incentive to focus attention on patients who have a high risk of hospitalization, but it could also cause practices to avoid accepting sicker patients or to reduce services to other patients.

7. Practice Revenues Would Be Unpredictable and Uncontrollable.  Because most of the costs in a primary care practice are fixed, a primary care practice needs to be able to predict how much it will receive each month in order to be sure it will have enough revenues to cover the costs of hiring additional personnel to deliver expanded services to its patients.  The current design for Primary Care First will make it impossible for primary care practices to predict how much they will be paid.  For example, in a small practice, one or two unexpected hospital admissions could result in a 7% reduction in payments.

8. Payment Complexity Would Increase and Administrative Burdens Would Remain High.  Although CMS says that Primary Care First will “allow clinicians to focus on caring for patients rather than their revenue cycle,” the practice’s “revenue cycle” will actually become more complex than it is today.

9. Most Primary Care Practices in the Country Will Be Unable to Participate.  Even if a primary care practice wants to participate in Primary Care First, it will not be able to do so if it is located in Alabama, Arizona, Connecticut, the District of Columbia, Georgia, Iowa, Idaho, Illinois, Indiana, Kansas (other than the Kansas City metro area), Kentucky (other than the Cincinnati metro area), Maryland, Minnesota, Mississippi, Missouri (other than the Kansas City area), Nevada, New Mexico, New York (other than the Buffalo and North Hudson regions), North Carolina, Pennsylvania (other than the Philadelphia Region), South Carolina, South Dakota, Texas, Utah, Vermont, Washington, West Virginia, Wisconsin, or Wyoming.  More than 40% of the Medicare beneficiaries in the country will not have an opportunity to receive care from a primary care practice participating in either Primary Care First or the Comprehensive Primary Care Plus initiatives.

The Direct Contracting Options Are Not Options for Most Primary Care Practices

Only large primary care practices with at least 5,000 beneficiaries would be eligible to participate in the new “Direct Contracting” options in the CMS Primary Cares Initiative.  Although they could receive a large, flexible “population-based payment” to deliver primary care services, they would also be required to pay CMS for 50%-100% of any increases in total Medicare spending for their patients beyond whatever benchmark spending level CMS establishes.  Even a small increase in total Medicare spending would represent a large proportion of a primary care practice’s revenue, putting the practice at risk of bankruptcy if it does not have large financial reserves available. 

Changing Primary Care First So It Provides the Support Small Primary Care Practices Need

Fortunately, it would be relatively easy for CMS to modify the details of the Primary Care First initiative to solve the problems described above.  Nine specific changes would enable CMS to create the kind of payment model that smaller primary care practices have been seeking.  These changes are based on the two primary care payment models recommended by the Physician-Focused Payment Model Technical Advisory Committee (PTAC) that were developed by the American Academy of Family Physicians (AAFP) and Jean Antonucci, MD (a solo primary care physician practicing in rural Maine).

1.  Pay practices with a monthly per-patient payment in place of all fees for office visits. 

2.  Pay a higher monthly amount for a patient who has greater needs. 

3.  Set monthly payment amounts at levels adequate to support high-quality primary care.  The monthly payment under Primary Care First should be no less than $65 per patient in 2020, and ideally even higher.  The monthly payment amounts for patients with more complex needs should be higher than this average amount.

4.  Begin paying monthly payments immediately when a patient enrolls for care in the primary care practice.  CMS already does this for Chronic Care Management (CCM) payments under the Medicare Physician Fee Schedule – as soon as a patient consents to receive CCM services, the physician practice can begin delivering the services and receiving payment for them.  A similar approach should be used in Primary Care First.

5.  Create billing codes so that primary care practices can classify patients appropriately and receive timely monthly payments for each patient.

6.  Evaluate performance using patient-centered outcome measures with achievable performance targets. 

7.  Establish performance-based rewards and penalties that create manageable levels of financial risk for small primary care practices.  The total adjustment to the monthly payment based on all aspects of performance should be no more than 15% of the monthly payment amounts, which in turn would be much higher than the payments primary care practices receive today.  Moreover, if every primary care practice performs well, every primary care practice should be rewarded for doing so, rather than limiting rewards to a small subgroup of practices. 

8.  Create a complementary payment model to support home-based palliative care for seriously ill patients.  Patients with serious illness need home-based palliative care, not just primary care.  Small primary care practices don’t have enough patients with serious illness to enable them to deliver palliative care services cost-effectively, and the payment amounts in the Primary Care First option for Seriously Ill Patients are far below what is needed to support palliative care services alone much less primary care, too.  In addition to an improved Primary Care First program for primary care practices, CMS should implement the payment models for palliative care that were developed by the American Academy of Hospice and Palliative Medicine (AAHPM) and the Coalition to Transform Advanced Care (CTAC), both of which were recommended for implementation by PTAC over a year ago. 

9.  Allow primary care practices in all parts of the country to participate in the revised Primary Care First program.  Every Medicare beneficiary deserves to receive high quality primary care services, and every beneficiary with a serious illness deserves to receive palliative care services.  Primary Care First should be expanded so that every primary care practice in every state has the opportunity to participate, and so that palliative care services can be delivered in every community.

More details on all of these changes are in The Problems With Primary Care First and How to Fix Them.

The Goal of Primary Care is to Improve Patients’ Health, Not Just to Save Money for Medicare

Higher and more flexible payments for primary care will enable primary care services to not only stay afloat but to deliver better services to their patients.  This will result in fewer avoidable hospitalizations, fewer unnecessary tests, and fewer inappropriate referrals to specialists, which in turn will produce significant savings for Medicare on these types of services.  However, it may be unrealistic to expect that these savings will be sufficient to fully offset the cost of the higher payments needed to adequately support primary care, much less to achieve net savings overall, during the 5-year time period typically used in CMS evaluations.  Placing more financial risk on primary care practices is more likely to accelerate the loss of primary care providers than to achieve greater savings for Medicare. 

Because of the need to take a longer-term view of the value of primary care, the Center for Medicare and Medicaid Innovation (CMMI) may not be the appropriate mechanism for successfully addressing the problems facing primary care.  It seems increasingly likely that Congressional action will be needed to create a truly effective primary care payment system.  The biggest benefits of primary care will be seen beyond the five-year time horizon used in CMMI demonstration projects, through slowing the progression of chronic disease and even preventing some diseases from occurring at all, not just trying to avoid hospitalizations for those who already have such conditions.  Only Congress can authorize making investments designed to achieve these longer-run benefits.  Consequently, in addition to revising Primary Care First to make it as successful as possible within current statutory constraints, CMS should ask Congress for the authority to create the kind of primary care payment system that the country truly needs.

NOTE: This post was authored by Harold Miller. Although Miller is a member of the Physician-Focused Payment Model Technical Advisory Committee (PTAC), the comments in this post are his personal opinions, and are not intended to represent the position of the PTAC as a whole or of other individual members. PTAC’s statutory charge is limited to reviewing proposals for payment models that are submitted to it by stakeholders, and it has no role in advising HHS or CMS other than submitting comments and recommendations on the proposals it receives.


 Comments (0)

Tuesday, January 22, 2019

Why CMS Alternative Payment Models Are Failing

The reason CMS Alternative Payment Models (APMs) are failing to significantly reduce spending in Medicare is that none of the major APMs meets the majority of the eight criteria for a successful APM.  The Problems with Medicare’s Alternative Payment Models and How to Fix Them details the serious weaknesses in the six major Advanced APMs that have been implemented by CMS and the CMS Innovation Center:

  • Bundled Payments for Care Improvement Advanced (BPCI-A);
  • Comprehensive Care for Joint Replacement (CJR);
  • Comprehensive End Stage Renal Disease Care (CEC);
  • Comprehensive Primary Care Plus (CPC+);
  • The newly adopted Basic Track Level E and Enhanced Track in the Medicare Shared Savings Program (MSSP); and
  • The Oncology Care Model (OCM).

As shown in the chart, not only do these APMs fail to meet the majority of criteria for a successful APM, most of the APMs fail to correct any of the problems with fee-for-service payment.  The reason is simple – most CMS APMs don’t actually change the way physicians, hospitals, and other healthcare providers are paid.  Most of the APMs don’t pay for high-value services needed to improve care nor do they ensure the amounts paid for key services match the costs of delivering those services.  Only CPC+ and OCM provide new payments to physicians so they can provide different services than they do today. 

What is even more problematic is that the spending accountability mechanisms included in most CMS APMs penalize healthcare providers for treating sicker patients and patients who need more expensive treatments, while providing no assurance to individual patients that necessary services won’t be withheld in order to meet spending targets.  Although all of the APMs include quality measures, the bonuses and penalties under the APMs are based on the average quality of care for all patients, not the quality of care for individual patients.  Moreover, the risk adjustment systems used in the APMs fail to recognize some of the most important reasons why patients may need more services or more expensive services to successfully treat their health problems.  As a result, under most of the APMs, physicians and hospitals could actually receive financial bonuses if they give some patients less treatment than they need, and providers could be penalized when they provide the most effective care for complex patients. 

CMS can do better – much better.  The Problems with Medicare’s Alternative Payment Models and How to Fix Them describes two different APMs designed to improve care for patients with chronic conditions.  Unlike the existing CMS APMs, each of these APMs would meet all eight of the criteria for a successful APM.

Moreover, these APMs would address a major opportunity for saving money and improving care for Medicare beneficiaries that CMS has completely ignored – providing the resources that teams of specialists and primary care physicians need to provide high-quality care to patients with chronic conditions.  Patients with asthma, chronic kidney disease, COPD, diabetes, heart failure, inflammatory bowel disease, and other chronic conditions are often hospitalized for exacerbations of their condition that could have been avoided through better care management services, in-home services, and other services, but these services aren’t paid for adequately or at all under current Medicare payment systems.  Yet despite the fact that a large percentage of Medicare beneficiaries have one or more of these chronic conditions and there are opportunities for significant savings by improving their care, CMS has not implemented any APMs specifically focused on helping patients with these conditions avoid hospitalizations.  None of the existing CMS APMs change payments for the specialists who typically manage or co-manage care for the subset of patients who are most at risk of complications, even though these physicians are often in the best position to help such patients avoid hospitalizations. 

Alternative Payment Models hold the potential for accelerating progress toward more affordable and higher-quality care, but only if they are designed in the right way and only if they are focused on opportunities to reduce spending without harming patients.  Faster progress in developing and implementing these types of APMs needs to be a national priority.

 , , ,

 Comments (0)

Friday, January 04, 2019

Does Medicare Need Mandatory Payment Models?

The Failure of Current Alternative Payment Models

Despite high hopes that Alternative Payment Models (APMs) would achieve significant savings for payers and better quality care for patients, most of the APMs implemented to date have had disappointing results. 

The largest APM in the nation is the Medicare Shared Savings Program (MSSP).  Debates among statisticians about how much savings have been achieved by Accountable Care Organizations in the MSSP have obscured the fact that even the most optimistic analyses show net savings of only $29 per beneficiary per year (one-quarter of one percent) during the first four years of the program (2013-16), and savings in 2017 were only slightly higher. 

The poor performance of current APMs is not surprising because they do not actually change the fee-for-service payment system; they merely pay bonuses if providers can reduce spending and/or impose penalties if spending increases.  However, because there are no changes in the way physicians, hospitals, and other providers are paid, they cannot deliver many kinds of high-value services that could reduce spending without harming patients.  Moreover, most APMs require that individual providers reduce total spending on their patients, even though no individual provider can control all of the services a patient receives or all of the factors affecting the costs of those services.

The Move to Higher-Risk APMs

Unfortunately, rather than creating better APMs, the Centers for Medicare and Medicaid Services (CMS) and other payers claim that the solutions are to continue using the same approaches, but with higher levels of financial risk for providers, or to create “population-based payments” in which healthcare providers would be expected to deliver all of the services a patient needs for a fixed monthly or annual payment.  CMS has already announced plans to phase out the “upside only” Medicare Shared Savings Program, and the Center for Medicare and Medicaid Innovation (CMMI) has indicated that most new APMs it tests will require high levels of financial risk.

Because of the problematic way that most APMs have been structured, many providers have been unwilling to participate.  Increasing the level of risk in APMs could result in even fewer providers participating, which in turn would reduce the total savings that can be achieved.  In addition, because participation in APMs has been voluntary, evaluations of the APMs will be biased if the initial participants are not representative of all providers who would ultimately participate.  

Concerns about these issues have led to calls for CMS to create “mandatory” payment models. 

What is a “Mandatory” Payment Model?

The term “mandatory payment model” is misleading because the intention is typically not to require every provider in the country to participate in a particular alternative payment model.  Instead, the goal is to make participation in an APM mandatory for some providers in order to “test” the APM, i.e. to evaluate its effectiveness in reducing spending and/or improving quality.  

In an effort to mimic a randomized controlled trial, a randomly-chosen subset of providers would be paid under the APM (whether they want to participate or not), while other providers (or a random subset of other providers) would be prohibited from participating in the APM (even if they would like to be paid under the APM).  CMS is using one version of this approach in the Comprehensive Care for Joint Replacement (CJR) APM.  CMS selected 196 of the 380 metropolitan statistical areas (MSAs) in the country, randomly chose 75 of those MSAs, and required all hospitals in 67 of the MSAs to participate in the CJR APM for two years.  Hospitals in the rest of the country were not permitted to participate in the CJR APM.  After the first two years of the program, the mandate was removed entirely in 34 of the MSAs and for small and rural hospitals in all 67, but about one-fourth of those hospitals elected to continue participating voluntarily.  

The Problems with Mandatory APMs

There are a number of serious problems with the “mandatory model” approach to APM testing:

  • There is no guarantee that the APM will achieve the desired results or that it will have no undesirable impacts, otherwise there would be no need to test it.  APMs designed to reduce spending, particularly those involving significant financial risk, have the potential to harm patients as well as providers. It is inappropriate to require providers and their patients to participate in an APM that could be harmful before any testing or evaluation has been completed. 
  • Conversely, if the potential benefits of the APM are sufficiently great to justify requiring some providers to participate, it is inappropriate to prevent other providers from voluntarily participating, particularly if their patients could benefit from the type of care that could be delivered with support from the APM.
  • It is well-known that there are significant differences in the health status of patients across the country and significant differences in practice patterns of providers both across and within geographic regions.  Randomization only by region without randomization of providers or patients within regions would only have a limited ability to separate the effects of the APM from the influence of inter-regional and intra-regional differences in spending and quality.
  • Requiring every provider in a geographic region to participate in a payment model does not preclude patients from traveling to a different region for services, either to avoid the effects of the APM in their home region or to obtain whatever benefits are expected under the APM from a provider in a participating region.  For example, there is anecdotal evidence that in regions where CJR is mandatory, higher-risk patients are being forced to travel to non-participating regions in order to receive hip or knee surgery.  This “cherry-picking” and “lemon-dropping” could reduce the reliability of any evaluation results.
  • Because the mandate may only last until the evaluation is completed, providers who would be unwilling to participate voluntarily may also be unwilling to make the changes in care delivery necessary to achieve success while the test is underway.  Consequently, the evaluation of a temporary mandatory test would not necessarily show what the impacts of a permanent mandatory program would be.  Indeed, if a subset of providers did not want the APM to be universally mandated, they would have an incentive not to produce successful results during the demonstration.

A Better Approach:
Well-Designed APMs That Attract Patients and Providers

Support for mandatory models rests heavily on three false premises.  False Premise #1 is that a voluntary approach will not provide an accurate indication of the true impact of an APM.  However, there is no evidence for this; in fact, the evaluations of the mandatory CJR APM and the voluntary Bundled Payments for Care Improvement (BPCI) APM found similar impacts on savings and quality for hip and knee replacement surgeries. 

False Premise #2 is that when providers are unwilling to participate in an APM, the only way to overcome that is to require them to participate.  However, there is a better alternative — redesigning the APM so that providers and patients will want to participate.  

Randomized controlled trials (RCTs) for drugs, medical devices, and procedures do not mandate that either patients or providers participate. RCTs recruit volunteers based on the potential benefits of a new therapy or the opportunity to achieve similar outcomes at lower cost. 

The reason providers have refused to participate in CMS APMs is the poor design of those APMs, not unwillingness to move away from fee-for-service payment.  The best evidence of this is the dozens of APM proposals that physicians have submitted to the Physician-Focused Payment Model Technical Advisory Committee (PTAC).  Over the past two years, PTAC has recommended 14 APMs that physicians want to participate in.  But CMS has refused to implement any of them. 

Instead, CMS continues to design APMs that place providers at financial risk for aspects of spending and quality they cannot control, while failing to give providers the resources they need to improve patient outcomes and reduce Medicare spending.  There is no evidence that simply increasing the financial risk in CMS APMs would result in greater savings.  Moreover, transferring significant financial risk to providers can have undesirable results, including loss of access to services for higher-need patients, higher prices due to consolidation of providers, and lower quality of care.   

A Better Approach:
Limited Scale Testing Followed by Phased Expansion

False Premise #3 is that a new APM must be immediately tested with a large number of providers in order to determine whether it works.  Yet this is the exact opposite of the approach used for clinical trials.  Large, randomized trials are Phase III, not Phase I.  The drug, device, or procedure is first tested with a small number of volunteers to ensure it is safe, then it is tested with a larger group of volunteers to make a preliminary assessment of its efficacy, and only then is a large scale trial contemplated.  Positive results from the initial phases encourage patients and healthcare providers to voluntarily participate in the larger-scale tests.

The mandatory APM approach is also the exact opposite of how every other industry develops new products and services.  New products do not immediately jump from the design table to large-scale production.  Businesses create one or more prototypes and then test them on a limited scale to identify problems and opportunities for improvement.  The design is then revised before broader production begins.  In addition, product designs continue to be refined and new versions of products are created whenever the benefits for consumers are expected to outweigh the costs of making the changes.

A similar “beta testing” process is needed for APMs.  The more innovative the APM – i.e., the more that it differs from the current payment system – the more likely there will be a need for initial beta testing and potentially for additional rounds of refinement after the APM is implemented more widely.  Positive results from the beta testing phase will encourage more patients and providers to participate in larger tests, and if results continue to be positive, they will pave the way for broad implementation.  This is why PTAC has recommended beginning with limited-scale testing for most of the APMs it has favorably reviewed.

Should Successful APMs Be Mandated?

If testing of an APM is successful, it will likely be appropriate to make it permanent.  However, participation can still be voluntary.  Successful APMs will not achieve savings simply because they pay in a different way or because they create an “incentive” to spend less; they will achieve savings by removing specific payment-related barriers to changing care delivery that will impact specific opportunities for improvement.  Both the opportunities to achieve savings and the barriers to pursuing those opportunities will differ from community to community.  This means that “one size fits all” APMs will be less likely to achieve the full amount of savings that are possible nationally or to provide better care to patients in all parts of the country than a more customized approach.  Mandating a single approach may lower administrative costs for Medicare and other payers, but it will also likely reduce the amount of savings even more.  Multiple, voluntary APMs will allow the fastest progress toward higher quality, more affordable healthcare.

(Download Does Medicare Need Mandatory Payment Models here.)

 , ,

 Comments (0)

Monday, December 17, 2018

How Can You Tell a Good Alternative Payment Model From a Bad One?

Most current Alternative Payment Models have failed to achieve significant savings or improvements in quality because they are badly designed – they don’t correct the problems with current fee-for-service payment systems, and they have the potential to harm vulnerable patients and force small healthcare providers out of business. 

How can you tell a good APM from a bad APM?  Here are the eight criteria you should use to evaluate whether a proposed APM is likely to be successful:

  1. Does the APM pay for the high-value services needed to improve patient care?  To be successful, an APM can and must make any changes needed in the way physicians, hospitals, and other healthcare providers are paid so they are able to deliver high-value services that will improve outcomes and reduce spending.  Most current APMs do not make any changes in the basic ways that providers are paid, but merely provide “incentives” to reduce spending or improve quality.
  2. Does the APM align the payment amount with the cost of delivering high-quality care?  To be successful, an APM must align payments with the actual costs of delivering high-quality services, particularly when the volume of services is reduced.  Many current APMs actually widen the gap between payments and costs rather than narrowing it. 
  3. Does the APM assure each patient they will receive appropriate, high-quality care?  In order to protect patients, an APM should be designed with patient-level quality standards and targets that assure each patient they will receive high-quality care and achieve good outcomes.  Most current APMs only assess whether quality has changed on average for a group of patients, not whether it has improved or worsened for individual patients. 
  4. Does the APM make the cost of diagnosing or treating a health condition more predictable and comparable?  An APM can and should specify in advance the amount that will need to be paid for treatment of a particular condition or combination of conditions so that patients can compare the costs of care across providers.  Many current APMs do not set spending targets until after services are already delivered, and most do not even make final determinations as to which patients are eligible for the APM until after services are delivered, making it impossible for a patient or their payer to know in advance how much will be spent on the patient’s care.
  5. Will a provider only be paid under the APM if a patient receives services?  Although there are clearly serious problems with the quality and cost of the services delivered under fee-for-service payment, fee-for-service at least gives patients and payers the confidence that they will only pay something if they receive something in return.  Under many “population-based payment” APMs, providers are paid even if they do nothing for patients. 
  6. Are payments under the APM higher for patients who need more services?  Although fee-for-service payment is criticized for rewarding “volume over value,” any payment system that doesn’t adequately support a higher volume of services when patients need more services can result in worse outcomes for patients and higher long-run costs.  Many APMs fail to adjust payments for important differences in patients that require more services or different types of services.
  7. Is a provider’s payment under the APM based on things the provider can control?  Although fee-for-service payment fails to hold physician, hospitals, and other healthcare providers accountable for problems they caused or could have prevented, it also does not penalize them for things outside of their control.  Many current APMs go too far in the opposite direction – placing healthcare providers at financial risk for the total cost of care even though they can only control or influence a small part of it. 
  8. Will a provider know how much they will be paid under the APM before delivering services?  Under fee-for-service payment, physicians, hospitals, and other providers know exactly what they will be paid for delivering a service before they deliver that service, so they can determine whether they are likely to receive sufficient revenue to cover their costs before they incur those costs.  Under many APMs, it is impossible for the participating providers to predict how much they will be paid for the services they will deliver, and they may not know for sure how much they will receive until many months after the services are actually delivered. 

As the chart below shows, the shared savings/shared risk and “population-based payment” designs used by most current APMs fail to meet the majority of these criteria.  (You can download a copy of the Criteria chart here.) 

Fortunately, there are ways to design APMs so they meet all eight of the criteria.  Four general designs will likely meet the need for an APM in most circumstances:

Option 1A: Accountable Payment for Service.  Under this APM design, a provider receives a new or revised payment for delivering a specific service to patients, and the payment is reduced if targets for spending on specific services and for performance on quality measures are not achieved.

Option 1B: Accountable Bundled Payment.  Under this design, a provider or team of providers receives a bundled payment to enable delivery of a group of services to patients or to treat a particular condition, and the payment is reduced if targets for spending on specific services and performance on quality measures are not achieved.

Option 2: Outcome-Based Payment.  Under this design, a provider is paid for a service or group of services only if standards or targets for quality and spending are achieved.

Option 3: Bundled/Warrantied Payment.  Under this design, a provider or team of providers receives a bundled payment to deliver a group of services to patients, and the provider team is responsible for using the payment to cover the costs of necessary services and also to pay for avoidable services or services needed to address complications of treatment.

The components of these four APMs are summarized in the Framework for Alternative Payment Models that can be downloaded here

The details on how to design each of these Alternative Payment Models are available in CHQPR’s report How to Create an Alternative Payment Model.


 Comments (0)

Sunday, October 28, 2018

Saving Primary Care Before It’s Too Late

Talk to almost any primary care physician these days and what you’ll hear should frighten you. Many PCPs are actively looking for ways to get out of primary care, and there’s no one to replace them because medical students don’t want to go into primary care. One well-known reason for this is that PCPs are being forced to spend more time dealing with the administrative burdens of dysfunctional EHRs, measurement programs, and health plan rules than taking care of patients.

But there’s another reason that nobody is talking about, and it may be an even bigger threat to the future of primary care.

The Hidden Flaw in “Value-Based Payment” Programs

Health plans and public programs simply aren’t paying enough to cover the costs of good primary care. Payments to PCPs from both Medicare and private payers have fallen far behind inflation, and so-called “value-based payments” do more harm than good when the payments aren’t enough to keep the lights on. Large health systems can afford to subsidize these losses, but independent primary care physicians can’t. They will simply be forced out of business, and that will be devastating for many small and rural communities.

You know it’s bad when primary care practices have to ask for donations in order to keep treating patients. The residents of Hilo, Hawaii created a fundraising site to try and prevent a young primary care physician and her practice from going bankrupt because of the low payments from the big health insurance company in Hawaii. You can see the plea for support here:

This young physician – Michelle Mitchell – has worked hard to create the kind of primary care practice we need everywhere in America. And she’s trying to do it in a small rural community with many poor people and a severe physician shortage. Hilo is not the wealthy “resort” part of Hawaii; on top of problems of poverty, homelessness, and unemployment, it’s been in the news over the past year because it has an active volcano destroying homes and polluting the air, and it was hit with a severe hurricane. It’s the kind of place where health plans should be doing everything in their power not only to keep primary care practices open but to provide extra support so they can deliver true primary care medical home services to their patients. But unfortunately, that’s not happening. You can read Dr. Mitchell’s story here:

There’s an important lesson here for many current “value-based payment” initiatives. Rather than working with the physicians in the community to figure out what kind of payment system would actually support good primary care in rural communities, the health insurance company in Hawaii hired a group of behavioral economics theorists at a large urban medical center to tell them how PCPs should be paid. They published a journal article touting how they designed incentives to achieve savings for the health plan and incentives to improve quality.

But it’s clear from the article that the consultants and health plan never even tried to answer a very fundamental question – what does it actually cost to deliver good primary care in Hawaii and would the payment system they designed cover those costs? You can’t “incentivize” a PCP to deliver higher-value care if the PCP is no longer practicing at all because they defaulted on their loans or mortgage or failed to make payroll.

How should the payment system have been designed to support small rural practices? Dr. Mitchell gave her recommendations at the National ACO, Bundled Payment, and MACRA Summit in Washington DC in June. She described her experience participating in both the CMS Comprehensive Primary Care Plus program and the payment system in Hawaii, and she showed how those programs increased her costs and failed to provide enough revenue to cover the costs, leaving her practice worse off financially than it was before. You can see her recommendations for what is needed to make high-quality primary care sustainable here.

Saving Primary Care Before It’s Gone

There’s been a lot of talk for many years about the need to improve payment for primary care practices, but there hasn’t been much action, particularly at the federal level. Medicare still doesn’t have a primary care medical home program available for PCPs in most of the country. The CMS Comprehensive Primary Care Plus initiative is only operating in 18 regions of the country, and even in those regions, only a subset of PCPs are able to participate. Primary care practices in Alabama, Alaska, Arizona, California, Connecticut, Delaware, Florida, Georgia, Idaho, Illinois, Indiana, Iowa, Maine, Massachusetts, Minnesota, Mississippi, Nevada, New Hampshire, New Mexico, North Carolina, South Carolina, South Dakota, Texas, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, Wyoming and in most of Kansas, Kentucky, Missouri, New York, and Pennsylvania are still stuck in the traditional Medicare fee for service payment system.

In December 2017, the Physician-Focused Payment Model Technical Advisory Committee (PTAC) recommended that the Secretary of Health and Human Services begin testing a new primary care payment model recommended by the American Academy of Family Physicians so that more primary care physicians would have the opportunity to be paid in better ways. PTAC specifically recommended that the AAFP model be tested in a way that would “facilitate rapid implementation on a broad scale,” and it emphasized the urgent need to preserve and strengthen primary care.

Ten months has now passed since PTAC recommended a new primary care payment model, and nothing has happened.

Rather than waiting years for CMS to design the “perfect” payment model and then waiting many more years while it’s tried in a small number of places, it’s time for a different approach. CMS and other payers should:

  • Try multiple payment models so we can learn which approaches work better in which places. It’s unlikely there will ever be a one-size-fits-all model that will work well in every community and every type of practice, so we should stop trying to develop one ideal model.
  • Let the physicians design the payment models in ways they believe will work. The current top-down, payer-driven approach to alternative payment models hasn’t succeeded in improving quality or reducing costs, so it’s time to try a bottom-up approach instead. Physicians have shown they will respond if they are given the opportunity. For example, Jean Antonucci, MD, a solo primary care physician in a rural part of Maine, developed an innovative primary care payment model that is simpler than other approaches and that uses patient-reported outcomes to measure quality and stratify payments. PTAC reviewed her proposal in September and recommended that CMS test it in addition to the proposal developed by the AAFP.
  • Move quickly to try new approaches. Every day that goes by waiting for the ideal model to be developed is a day that a patient may not get the care they need and a day that a primary care practice comes closer to closing.

Although we’re spending too much overall on health care, we’re spending too little on primary care – estimates indicate that primary care represents less than 8% of total healthcare spending. If we don’t invest in better primary care, we’re not likely to be successful in reducing the 92% of spending where the opportunities for savings exist. The net cost of a 25% increase in payments to PCPs is zero if the reductions in avoidable hospitalizations and unnecessary tests and procedures from better primary care reduces other spending by a mere 2%.

The question is not whether we can afford to improve primary care, but whether we can afford not to. Will there be any primary care practices left by the time we finally decide to pay them the right way?


 Comments (3)

Thursday, August 30, 2018

Was the Medicare Shared Savings Program Successful in 2017?

CMS released the 2017 results for the Medicare Shared Savings Program (MSSP) on August 30, 2018. Here is what they show:

  • The 472 Accountable Care Organizations (ACOs) in the MSSP spent nearly $1.1 billion less than their “benchmark” spending levels (the amounts CMS projected that it would spend on the ACOs’ patients in the absence of ACO actions).
  • CMS gave $799 million of the savings back to 162 of the ACOs in shared savings bonuses. 16 Track 2 & 3 ACOs paid penalties to CMS totaling $57 million.
  • The net result was that CMS saved $313.7 million on the MSSP.

Is $313.7 million a lot of savings? Hardly:

  • The savings amounted to only $36 for each of the nearly 9 million beneficiaries in the ACO program – that’s what the ACOs would save if half of their patients made one fewer visit to the doctor each year.
  • The savings amounted to only 0.33% of the total spending of $95 billion in the 472 ACOs – one-third of one percent.
  • The $314 million in net savings combined with the total of $384 in net losses in the first four years of the program means that CMS has yet to generate a net benefit for the Medicare program after five years of trying.

Did the ACOs that took downside risk produce more savings? No, they actually saved less:

  • The 39 “downside risk” ACOs only saved an average of $27 per beneficiary (0.24%).
  • The 433 “upside-only” ACOs saved $37 per beneficiary (0.34%).
  • The upside only ACOs saved 36% more per beneficiary than the two-sided risk ACOs.
  • Only 59% of the downside risk ACOs reduced Medicare spending, and 60% of the upside-only ACOs reduced Medicare spending. 41% of the downside risk ACOs actually increased Medicare spending.
  • The downside risk ACOs spent $254 more in total per beneficiary ($10,933) than the upside-only ACOs did ($10,679) even after they “saved” money for Medicare.
  • Although the MSSP program didn’t save very much overall, 93% of the savings came from the upside only Track 1 ACOs.

The experience in 2017 indicates that forcing all ACOs to take downside risk would likely produce even less Medicare savings, not more. The path to savings isn’t more risk, but a completely different approach. How to Fix the Medicare Shared Savings Program explains why shared savings, shared risk, and other risk-based population payment systems are unlikely to ever result in significant savings for the Medicare program, and it describes the kind of patient-centered payment system that CMS and other payers should be pursuing instead.


 Comments (0)

Monday, March 19, 2018

How to Create More Successful Alternative Payment Models More Quickly

There is broad consensus that the fee-for-service payment systems currently used in Medicare create significant barriers to delivering higher-quality care to Medicare beneficiaries at a more affordable cost. There is also growing concern that rather than solving the problems with fee-for-service payments, Medicare’s pay-for-performance programs are making things worse. Properly-designed Alternative Payment Models (APMs) are needed to solve the problems with both current fee-for-service payment systems and pay-for-performance programs.

Problems With CMS’s Current Approach to Creating Alternative Payment Models

In 2010, Congress created the Center for Medicare and Medicaid Innovation (CMMI) and provided it with $10 billion in funding over a ten year period in order to create Alternative Payment Models. However, after more than seven years and more than $5 billion in spending, only six payment models created by CMMI meet the APM requirements established by Congress, and only one of the models tested by CMMI has been expanded nationally. The slow progress in implementing successful Alternative Payment Models means that each year, millions of Medicare beneficiaries are being denied the opportunity to receive higher-quality care and the Medicare program is spending billions of dollars more than is necessary.

The lack of progress is due to several problems with the approach CMMI has used for developing and testing Alternative Payment Models:

1.   Most of the Models Being Tested Don’t Solve the Problems with Fee-for-Service Payment

Most of the large Alternative Payment Models created by CMMI – the Bundled Payments for Care Improvement (BPCI) model, the Comprehensive Care for Joint Replacement Model, the Comprehensive ESRD Model, the NextGen ACO Model, and the Pioneer ACO Model – follow the same basic formula:

  • No changes are made in the current fee for service structure. In most CMMI APMs, physicians, hospitals, and other providers continue to be paid under standard Medicare payment systems. No additional payments are made for any additional services, even if those additional services would be necessary or highly desirable in improving patient care and reducing avoidable spending.
  • Additional payments are dependent on achieving “shared savings.” A year or more after services are delivered, the physicians, hospitals, or other providers in the APM may receive an additional payment (or be required to repay some of the payments they have already received) based on whether CMS determines that it spent less than it otherwise would have. This approach is essentially the same as CMS’s pay-for-performance programs, except that the bonuses and penalties are proportional to the amount of money saved. This means that providers who already have high levels of performance receive no additional resources to sustain their operations, while providers who have had high rates of complications or who have overused expensive services can receive large bonuses for addressing those problems.

Adding “shared savings” on top of the current fee-for-service payment system does not solve the barriers in the current payment system, since participating providers may or may not receive adequate or timely funding to support new high-value services that would benefit their patients. An even greater concern is that a shared savings program can financially reward a healthcare provider for failing to order or deliver a costly service that a patient needs, since the provider could receive a portion of the savings when fewer services are delivered.

Moreover, the shared savings payment approach is already being used extensively for Accountable Care Organizations as part of the Medicare Shared Savings Program (MSSP). Rather than trying to find significantly different approaches to Alternative Payment Models, CMMI has focused a large portion of its resources on developing and testing variations on the MSSP model. In fact, during the first six years of CMMI’s existence, it spent over $600 million on four different variations on the MSSP ACO model (Pioneer ACOs, NextGen ACOs, the ACO Investment Model, and the Advance Payment ACO model).

Only two current CMMI Alternative Payment Models – the Comprehensive Primary Care Plus Model and the Oncology Care Model – provide significant new, upfront payments that are specifically designed to address barriers in the current fee-for-service payment system. However, continued payments under the Oncology Care Model are contingent on the oncology practice achieving savings. Even though CMMI tried using the shared savings model in the Comprehensive Primary Care Initiative and concluded that it was not an effective mechanism of paying physician practices, it proceeded to use a similar approach to pay oncology practices as part of the Oncology Care Model.

2.   Too Few Physician-Focused Payment Models Are Being Tested

Not only do the APMs developed by CMMI fail to solve the problems with fee-for-service payment and create potentially problematic incentives to stint on care, most are designed in one or more ways that make it difficult for physician practices, particularly small physician practices and single-specialty practices, to participate.

  • No upfront payments for new or enhanced services. Although upside-only shared savings models are commonly portrayed as “risk-free” for providers, the fact is that if the providers want to change the way they deliver care, they incur significant financial risk, since they have to pay for any new or enhanced services themselves and then hope that they will qualify for a large enough shared savings payment to cover those unreimbursed costs. This is particularly difficult for small physician practices and hospitals.
  • Eligibility limited to hospitalized patients, patients receiving expensive procedures, and complex patients. Most of the CMMI models developed to date have focused on reducing spending after a patient has already been hospitalized (e.g., the Bundled Payments for Care Improvement initiative), reducing spending on patients who are receiving expensive outpatient procedures (e.g., the Oncology Care Model for chemotherapy and the Comprehensive ESRD Care model for dialysis), and reducing spending on complex patients with multiple health problems (e.g., Accountable Care Organizations). These models ignore the many opportunities that exist to reduce spending by preventing hospitalizations, using lower-cost alternatives to expensive procedures, and preventing the development and progression of chronic diseases. Moreover, by limiting alternative payment models to patients who receive expensive procedures, the CMMI models could unintentionally encourage delivery of unnecessary procedures.
  • Participation restricted to hospitals. Even though physician practices have been successful in managing bundled episode payments for orthopedic and cardiac procedures in the private sector and in the Bundled Payments for Care Improvement initiative, CMMI has only permitted hospitals to participate in the Comprehensive Care for Joint Replacement (CJR) program, and the proposed CMMI Episode Payment program for cardiac procedures would have been open only to hospitals.
  • Requirements for large numbers of patients. In many of its models, the minimum number of patients CMMI requires for participation is much higher than the number of patients that small physician practices care for.
  • Financial penalties for providers based on things they cannot control. In most CMMI payment models, participating providers are rewarded or penalized based on whether the total Medicare spending on their patients is lower than expected, even if the providers have no ability to control or even influence all of the services their patients receive. For example, in the Oncology Care Model, oncologists could be penalized for increases in spending due to increases in drug prices, treatments for injuries the patients receive in car accidents, or complications of treatment delivered by other physicians for health issues unrelated to their cancer, even though the oncologists could not reasonably be expected to control or even influence many of these cost drivers.
  • Failure to risk adjust spending and quality measures based on clinical characteristics of patients that affect costs and outcomes. A physician, hospital, or other healthcare provider can significantly influence the cost and quality of care through the decisions they make about how to treat patients and the way they deliver treatments, but they cannot control how sick or frail the patients who come to them for treatment are. The risk adjustment systems used in CMMI APMs fail to recognize many of the key patient characteristics that affect costs and outcomes, which means that providers who care for sicker or higher-risk patients can face financial penalties, which in turn could make it more difficult for such patients to obtain the care they need.

Despite the fact that most of the healthcare services received by Medicare beneficiaries are delivered by non-primary care specialists, the CMMI Alternative Payment Model portfolio is almost devoid of models specifically designed for such specialists. In 2014, CMMI issued a Request for Information asking for input on the creation of specialty-specific payment models , and it convened several Technical Expert Panels to explore payment models in specialties such as cardiology, gastroenterology, neurology, and oncology. However, in the three years since then, CMMI has implemented only one APM – the Oncology Care Model — that was specifically designed for participation by small, non-primary care specialty physician practices.

In 2015, Congress recognized that CMMI had done too little to create Alternative Payment Models that were specifically designed for physicians. In the Medicare Access and CHIP Reauthorization Act (MACRA), it specifically encouraged the development of physician-focused payment models (PFPMs) by:

  • requiring the Secretary of Health and Human Services to establish criteria for PFPMs, including models for specialist physicians;
  • authorizing individuals and stakeholder entities to submit proposals for PFPMs that meet these criteria;
  • creating the Physician-Focused Payment Model Technical Advisory Committee (PTAC) to review these proposals and to prepare comments and recommendations to the Secretary regarding whether the proposals meet the criteria; and
  • requiring the Secretary of Health and Human Services to publicly post a detailed response to the PTAC’s comments and recommendations.

In its initial year of operation, PTAC received dozens of proposals for physician-focused payment models, and in April 2017, only four months after receiving the initial proposals, the PTAC recommended two payment models for limited-scale testing. These payment models would have enabled a wide range of specialists to participate in APMs that would improve care for a much broader range of patients than CMMI models. Unfortunately, in the Secretary of HHS’s response to PTAC, CMMI indicated it was not willing to implement the models PTAC had recommended.  PTAC recommended additional payment models for implementation and testing in October 2017, but five months later, the Secretary of HHS had not responded to those recommendations.

3.   The CMMI Testing Process is Slow, Burdensome, Expensive, and Discourages Significant Innovation

When CMMI decides to pursue development and testing of an Alternative Payment Model, the process it uses is extremely long, complex, and resource-intensive. This not only slows down the process of testing and implementation but it reduces the number of models that CMMI can or will test. Although many proposals for innovative alternative payment models have been submitted to CMMI, most have not been implemented even after many months of discussion with CMMI staff and despite efforts to modify proposals to address concerns raised by CMMI. Once CMMI decides to pursue a payment demonstration, it typically takes 18-24 months or more from the time an initiative is first announced to the time when providers actually begin to receive different payments. Even if a payment model is succeeding and other providers would like to participate, the evaluation process will take 3-5 years to complete before a decision is made as to whether a payment model should be continued or expanded. As a result, under the current process for implementing APMs, it will take 6-8 years to make a desirable alternative payment model broadly available.

Providers report that the process of applying to participate in CMMI payment models is very burdensome. Providers are expected to complete lengthy application forms requiring submission of data and other information that is expensive and time-consuming to assemble, and applications may be rejected for failure to meet non-substantive requirements such as maximum page limits. Applicants may be required to respond within a few days to CMMI’s requests for more information, but the applicants receive no commitment from CMMI as to when it will make a decision regarding their application. This uncertainty makes it difficult for a provider organization to know whether and when to start preparing for participation; starting preparation too soon could mean significant financial losses if the applicant is not accepted, whereas waiting until an application is approved to begin implementation planning could make it difficult for the provider organization to generate savings and quality improvements in the timeframes required in the demonstration.

Once accepted into a CMMI APM, providers are required to assemble and submit large amounts of data and to participate in a variety of meetings; these administrative activities can involve significant costs for providers and/or take significant amounts of their time away from patient care. There is generally little or no compensation provided to practices to offset these costs, even though CMMI spends tens of millions of dollars to pay the consultants who review the information the providers submit and organize the meetings they attend. Many providers, particularly small providers, have decided not to even apply to participate in otherwise desirable CMMI programs and others have dropped out of the programs in the early phases solely or partly because of the cost and time burden of participating.

Providers who do participate in CMMI payment models are told they can only count on the new payments lasting for a few years; the payments will only be continued beyond that if an evaluation proves that the program has saved money for the Medicare program. While this might sound like a very prudent approach, it can have the perverse effect of reducing the chances of significant success. Physicians, hospitals, and other healthcare providers are unlikely to fundamentally change the way they deliver care in response to a payment change that may only last a few years, and it is impossible to measure longer-term impacts on outcomes during an evaluation period that lasts only a few years.

How CMS Can Implement More and Better Alternative Payment Models Faster

CMS and CMMI could dramatically accelerate the implementation of APMs through the following steps:

1.   Embrace a Bottom-Up Approach to Payment Innovation at CMS

As part of MACRA, Congress created a bottom-up approach that specifically welcomes APMs designed by physicians and other practitioners and that encourages development of APMs and delivery models that are feasible for small physician practices and small hospitals to implement. CMS should embrace the process that Congress has created and commit to quickly implement each of the physician-focused Alternative Payment Models that is recommended by the Physician-Focused Payment Model Technical Advisory Committee.

In addition, there are a number of communities across the country where physicians, hospitals, and other providers are working with patients, employers, health plans, and other purchasers to develop and implement alternative payment models to support high quality, more affordable care. If a group of providers and payers in a state or region have developed or implemented an innovative APM, CMS should agree to implement a similar approach to paying for the care of Medicare beneficiaries in that community so that the providers can have full multi-payer support.

2.   Create the Capacity at CMS and its MACs to Implement Bundled Payments and Other Alternative Payment Models

There is no single alternative payment model that will work for all types of patients and all types of healthcare providers. CMS and its Medicare Administrative Contractors (MACs) should quickly make any changes needed in Medicare claims payment and other administrative systems to support implementation of seven types of alternative payment model structures.

1. Payment for a High-Value Service.
2. Condition-Based Payment for Physician Services.
3. Multi-Physician Bundled Payment.
4. Physician-Facility Procedure Bundle.
5. Warrantied Payment for Physician Services.
6. Episode Payment for a Procedure.
7. Condition-Based Payment.

In addition, CMS should revise the definition of “financial risk” in the MACRA regulations for Advanced APMs to enable design of APMs that small physician practices can feasibly participate in. CMS should also designate any APM that is undergoing testing by the Innovation Center as an “Advanced” APM.

3. Use Limited Scale Testing to Accelerate Innovation

Fully specifying the parameters of an innovative Alternative Payment Model often requires information that can only be obtained from providers who are delivering services in a different way, but providers cannot deliver services in that way without having an alternative payment model to support them. Currently, CMMI will only test a payment model if it projects that the model will reduce Medicare spending. However, since it is impossible to confidently make such a projection without specifying the parameters of the model, CMMI’s current approach means that most innovative models will never be tested.

To address this problem, CMMI needs to create a process for “limited scale testing” of innovative alternative payment models. The following five-step process could be used for this:

1. Selection of pilot sites for limited-scale testing.
2. Implementation of the APM at the pilot sites with adjustment of the parameters based on experience.
3. Collection of data for setting APM parameters for future expansion.
4. Decision about broader-scale testing or implementation of a refined model.
5. Transition of pilot sites if a model is not continued.

4.   Create a Faster, More Efficient Approach for Implementing APMs

If CMMI continues to use its current process for testing and implementing alternative payment models in the future, it would take a decade before the majority of physicians in the country would have the ability to participate in an APM designed for the types of patients they care for. CMMI should completely redesign the processes it uses to test and implement alternative payment models in order to achieve the goals that are implicit in MACRA – every physician should have the opportunity to receive at least 25% of their revenues from alternative payment models in 2019, at least 50% of their revenues from APMs in 2021, and at least 75% of their revenues from APMs in 2023. To ensure that the MACRA goals are achieved, CMS should establish specific milestones that are designed to implement as many alternative payment models as possible and as quickly as possible.


 Comments (0)

Thursday, April 13, 2017

A Revolution in Payment Reform Has Started

The first-ever “physician-focused payment models” were approved this week by the federal Physician-Focused Payment Model Technical Advisory Committee (PTAC). Instead of the traditional “top-down” approach to payment reform, where Medicare and other payers design payment systems that often don’t work well for physicians or patients, PTAC was created by Congress to enable a “bottom-up” approach in which physicians can design payment models that support the kind of care they believe patients need.

Project Sonar

The first of the payment models approved by PTAC was developed by a gastroenterologist in Illinois, Dr. Lawrence Kosinski, to support better care for patients with inflammatory bowel disease (IBD). Thanks to the willingness of Blue Cross Blue Shield of Illinois to implement the model with commercially-insured patients, Kosinski has already been able to show how having a better payment system can both improve patient care and save money. Patients who have been part of the project wrote to the PTAC describing how much better their care has been, and the health plan has said that it is saving significant amounts of money because patients haven’t been hospitalized as often.

But in order for all IBD patients to benefit from this improved care, Medicare and other payers around the country also need to implement the payment changes. Moreover, although the care delivery approach has been primarily used for IBD patients so far, it has the potential to support better care for patients who have other kinds of chronic health problems that need to be monitored closely to avoid the need for hospitalization, but those patients won’t be able to benefit either unless appropriate changes are made in the payment system.

Dr. Kosinski calls what he’s doing “Project Sonar” because he needed a way to regularly “ping” his patients to find out if they were developing problems so he could intervene early rather than only finding out after a patient is so sick they have to be hospitalized. The problem is the current fee-for-service system used by Medicare and most payers doesn’t pay for proactive contacts with patients and other care management services, so a physician practice that tried to do this routinely wouldn’t be able to cover the costs. Medicare and commercial health plans pay for office visits with the physician, but only if the patient schedules an appointment; the payers don’t pay for contacting the patient by phone or email to determine whether the patient needs a visit or not. The result is that Medicare and health plans end up paying for a lot of emergency room visits and hospitalizations that could have been avoided. That’s called “penny wise and pound foolish.”

ACS-Brandeis Advanced Alternative Payment Model

The second payment model approved by PTAC is very different. It’s designed to pay a wide range of physicians more when they deliver services as efficiently and effectively as possible during “episodes” of care, rather than paying them more for delivering unnecessary services, which is what the current fee-for-service system does. Although Medicare and other payers have tried to reward efficiency through Accountable Care Organization programs, those programs haven’t been successful because they don’t provide a way to change payment for the individual physicians who are actually delivering care to patients. The payment model developed by the American College of Surgeons and Brandeis University focuses payment changes on the teams of physicians (the “Clinical Affinity Groups”) that deliver specific types of care to patients. Moreover, it ties payment to how well the patient’s specific health problems are addressed, not just on how an individual procedure is delivered. Almost all of Medicare’s current “episode” payment models are limited to hospitalizations, which means that a patient has to be hospitalized before they can benefit from better care delivery, and in most cases the episode paymemt models are focused on surgeries, which ignores patients who can be treated without surgery (e.g., using medications or physical therapy). At the same time, the ACS-Brandeis model ensures that when surgery or another type of procedure is needed, the payment for the physicians who deliver the procedure is coordinated with the payments for the physicians who manage the underlying condition both before and after the procedure is performed, rather than forcing physicians to argue about “who generated the savings.”

What’s Next

The PTAC recommended both Project Sonar and the ACS-Brandeis Advanced Alternative Payment Model for “limited scale testing.” The PTAC defined this as a specific recommendation category because in many cases, specific elements of a payment model need further refinement before the model can be implemented on a broad enough scale to fully evaluate its impact, but that refinement cannot be completed without actually putting the payment model into operation in at least a small number of physician practices. For example, in the case of Project Sonar, although the payment model has already been used successfully for commercially-insured patients with inflammatory bowel disease, it is not clear how key parameters, such as the payment amounts and the spending targets, should be set for Medicare patients who are more likely to have multiple health problems and for whom additional time and resources may be needed to support both proactive outreach and coordination with other physicians.  The only way to choose the right parameters is to put the payment model in place so the costs and savings can be determined.

The PTAC’s recommendations are just that – recommendations to the Secretary of Health and Human Services that the two payment models should be tested in Medicare. The decision to actually carry out that testing is made by the Secretary, not PTAC, and then the Centers for Medicare and Medicaid Services would have to carry out the work needed to actually begin paying practices differently. Hopefully, the Secretary will agree with the PTAC’s recommendation and CMS will quickly implement the limited scale testing process.

Health insurance is becoming more and more unaffordable every year and the rapid growth in Medicare spending threatens the long-term viability of the program. The only way to solve those problems is to deliver health care in different ways. It is clear that there are many opportunities to significantly reduce healthcare spending today without harming patients, but it is also clear that the current payment system is a major barrier to achieving those opportunities. When solutions to the barriers are developed, they need to be implemented as rapidly as possible. Moreover, the payment changes need to be implemented in collaboration with physicians to support truly win-win-win outcomes – better care for patients, lower spending for Medicare, and financial viability and professional satisfaction for physicians.

First Steps in a Truly Revolutionary Process

Project Sonar and the ACS-Brandeis Advanced Alternative Payment Model are just the first of what will hopefully be many new physician-focused payment models developed by physicians, not payers. Congress recognized that the current approach to developing alternative payment models wasn’t working well, and it created the Physician-Focused Payment Model Technical Advisory Committee so that physicians would have the opportunity to develop better approaches. When he was a member of Congress, Secretary Price supported creation of the PTAC, and he attended the PTAC meeting on Tuesday to say that he felt physician-focused payment is key to an accessible, affordable, high-quality healthcare system.

The very first payment models PTAC considered demonstrate the wisdom of this approach. The payment model proposals were developed by practicing physicians, who attended the meeting in person and described the problems they saw with the care patients were receiving today and the barriers to improvement created by the current payment system. Their proposals and presentations explained how care could be better and spending could be lower with the payment models and care changes they had designed. The physicians also demonstrated that doctors can actually be enthusiastic about implementing a well-designed alternative payment model; they don’t have to be forced into it or given “incentives” to use it.

Additional proposals for physician-focused payment models have already been submitted to PTAC and many more are in the pipeline. The PTAC is also revolutionary because all of these proposals are available to the public, and the public has an opportunity to participate in the PTAC review process from the very beginning. The PTAC only makes its decisions about a proposal after obtaining public input on the proposal when it is first submitted, and all of the PTAC’s deliberations about the proposal are conducted in a public meeting with opportunities for public comment on a draft report prepared by a subset of the PTAC members.

More information about PTAC and copies of all of the proposals and letters of intent that have been submitted to date are available on the PTAC website at . You can receive email updates on the PTAC work through the PTAC email listserv and you can also follow PTAC on Twitter at @PFPMTAC.


 Comments (0)

Newer Posts »

©2008-2011 Center for Healthcare Quality and Payment Reform. All rights reserved.
320 Ft. Duquesne Blvd., Suite 20-J - Pittsburgh, PA 15222 - (412) 803-3650 -