This article originally appeared in Law360 on July 19th, 2021. It is republished here with permission.
Alternative payment models, or APMs, that are designed to reshape our health care delivery and payment system have now existed for nearly a decade.
Since the passage of the Affordable Care Act, the U.S. Centers for Medicare and Medicaid Services has implemented the statutorily mandated Medicare Shared Savings Program and CMS' Center for Medicare and Medicaid Innovation has implemented and tested over 50 APMs.
The Medicare Payment Advisory Commission, or MedPAC, recently issued a report to Congress that included a chapter reviewing APMs and making recommendations concerning APMs. The principal recommendations are that the number of APMs should be more limited and that they be structured with less overlap in a manner to permit better evaluation of their success.
While the MedPAC recommendations are well reasoned, the report does not directly address what attributes of APMs have been successful or have failed, and which should be included or not included in APMs going forward.
In addition, after about a decade of testing various models, it would be appropriate to ask where the U.S. health care system is headed with APMs. MedPAC's recommendation to continue the testing with fewer models does not address that question.
MedPAC was established in the late 1990s to provide advice to Congress on the Medicare program. Among its activities, MedPAC issues two reports a year to Congress, which include recommendations concerning various aspects of the Medicare program, including Medicare Advantage. The report dated June 15, 2021, includes 10 chapters, each addressing a different topic and with different recommendations.1
One of the topics addressed in the report is an assessment of the various APMs that CMS and the Center for Medicare and Medicaid Innovation have implemented and tested since the passage of Affordable Care Act more than 10 years ago.2
The Center for Medicare and Medicaid Innovation was charged in the ACA to test innovative payment and service delivery models that will reduce government program spending while enhancing or preserving health care quality.
In addition, CMS implemented the Medicare Shared Savings Program, the largest APM, as required by the ACA, which had a similar goal.
Currently, in 2021 there are 12 APMs operating that offer 25 distinct tracks in which providers participate. Typically, a Center for Medicare and Medicaid Innovation-introduced model is designed to last for around five years, and, if it is successful, it may reappear in a revised version of the model.
Not all the various APMs have been successful, though MedPAC noted in the report that, overall, APMs may have had promising results or "sufficiently actionable lessons learned."
There have been modest gross savings to the Medicare programs in many of the models, and some have generated net savings to the Medicare program even after taking into account the payments under the model made to participating entities; others did not achieve Medicare savings. The report includes a useful chart showing the success of the APMs on a model-by-model basis.3
MedPAC in its report also discusses various barriers to the success of the various APMs, many of which were predicted.
The report also discusses an additional issue with the success of APMs, namely, that many APMs allow providers and beneficiaries to be in more than one APM at the same time which can lead to problematic interactive effects including a weakening of the incentives and making it more difficult to assess the impact of a particular APM.
While MedPAC commended CMS and the Center for Medicare and Medicaid Innovation for their initiative in developing and testing so many APMs, based on its assessment of the APMs, MedPAC recommends in the report that CMS "now take a more holistic approach that involves implementing a smaller, more harmonized portfolio of APMs that are designed to work together."
MedPAC believes that continuing to test many independent APMs is likely to "inhibit the ability of APMs to reach their full potential."
In the report, MedPAC suggests that CMS consider a few different approaches with the more limited number of APMs.
First, it suggests that CMS focus on "a single population-based model with different tracks by provider type or beneficiary population." Examples are separate tracks for integrated health systems, multispecialty physician practices, end-stage renal disease facilities or the like.
As a second approach, MedPAC suggests consideration of a geographic approach to testing models. Such an approach would lessen the detrimental impact of overlapping and competitive models.
As a third and final suggestion, MedPAC recommends having CMS encourage states to pursue waivers that allow them to offer a smaller set of state-specific versions of various APMs, such as the Vermont All-Payer Accountable Care Organization Initiative and the Maryland Primary Care Program.
MedPAC finds sufficient merit in APMs for the continued testing of APMs in coming years, albeit with the recommendation that CMS utilize a smaller number of them with certain direction in their focus.
MedPAC's recommendations are well reasoned and are based on identified results from the review and evaluation of the over 50 APMs that have been tested.
MedPAC, however, does not emphasize or evaluate in depth whether there are certain successful components of APMs that should be included in the smaller number of APMs that should continue to be tested. To those of us with active practices involving APMs, it has appeared, but not fully evaluated, that there are both recurring challenges for success as well as components that successful APMs have in common.
Both warrant further evaluation and could lead to helpful recommendations of the structuring of successful APMs as CMS moves forward.
For example, providers have been reluctant to participate in APMs in which they must take downside risk. Why is that? Is it because they do not have the tools in place to be successful? Or, perhaps it is just their aversion to taking risk until required to do so.
An analysis of the provider's motivation, perhaps through interviews with providers, could provide some helpful perspective on that question.
Further, in many of the APMs, providers continue to receive the great bulk of their compensation on a fee-for-service basis with a smaller percentage of their potential compensation being at risk or tied to the incentives built into the APM.
Given that, can we really expect providers to be driven by the incentives where they must cut back on their more predominant fee-for-service revenue to earn the incentive? The Medicare Shared Savings Program, for instance, provides that for each dollar of reduced revenue from Medicare, an accountable care organizations may earn an incentive of 50 cents.
It also appears that in successful APMs, such as the physician-led Medicare Shared Savings Program accountable care organizations and the Comprehensive Care for Joint Replacement model, clinicians have received financial incentives by reducing services provided by providers other than the clinicians themselves.
In the Comprehensive Care for Joint Replacement model, for example, much of the savings resulted from savings generated from utilization of post-acute services. In physician-led accountable care organizations, reducing hospital services may lead to additional payments to the participating physicians.
These examples and no doubt others like them warrant a further directed review and evaluation to assess the structure of APMs and how APMs should be designed to generate more savings.
Similarly, it would be helpful to evaluate whether mandatory or voluntary participation produces better results and whether there is a structure that does not disadvantage providers who have previously taken steps to be more efficient. This is especially important now, as CMS has indicated its desire to implement new mandatory models.
In addition, when APMs were first introduced, some predicted a transformation to pay for performance on a much faster pace. After about a decade, it is fair to ask, where is the experimentation with APMs ultimately headed? And, is it appropriate to move the health care delivery system to one that pays for performance?
MedPAC in its report does not address those significant questions, no doubt in part given the need for further evaluation. Rather, it concludes that more of the same is indicated, namely testing and analysis, albeit of more streamlined APMs.-----------------------------------------------------
1 Medicare Payment Advisory Commission, "Report to the Congress: Medicare and the Health Care Delivery System," June 15, 2021 (the "Report").
2 Chapter 2 of the Report, "Streamlining CMS's Portfolio of Alternative Payment Models."
3 See Report at pages 51-53.