This past January, Health and Human Services Secretary Burwell announced her department’s goal to tie 50 percent of traditional Medicare payments to alternative or value-based payment models over the next three years. More recently, CMS announced the agency’s newest Accountable Care Organization (ACO) demonstration, Next Generation, certified the expansion of the Pioneer ACO demonstration, and earlier this month reformed the Medicare Shared Saving Program (MSSP) via rule making. These actions, combined with recent passage of the so called “doc fix” legislation, titled the Medicare Access and CHIP Reauthorization Act (MACRA), clearly indicate the federal government and CMS are rapidly transitioning the Medicare program from paying for volume to paying for value.
While tying payment to value makes perfect sense, transforming the Medicare program without the evidence that explains how to do this does not. By confounding theory and practice, CMS has made pay for value, a means by which cost efficiency and clinical practice can be improved, an end in itself. The corollary, it appears, is that paying for value via risk contracting is a prerequisite. This explains why CMS made only limited improvements to Track 1 in its final MSSP rule despite the fact that Track 1 participants constitute essentially the entire Shared Savings Program.
What’s Currently Known About ACOs?
Pioneers
CMS’s initial ACO effort, the Pioneer ACO demonstration, began in 2012. Here’s what we know about the demo. To date, 13 of the 32 Pioneer providers have quit the program. Nine of these lost money in their first year. When Sharp Healthcare exited the demonstration last fall, its CEO stated the demo’s restrictive rules left Sharp “harmed,” adding further, “That’s basically the case for all California ACOs.”
Performance results have been solidly mixed. According to CMS’s evaluation contractor L&M Policy Research, 12 Pioneers had no statistically distinguishable savings over the first two performance years 2012 and 2013. Ten had statistically significant savings over two years and ten in only one year. Year-over-year savings fell off by over 60 percent. Seventy percent of all savings were achieved by three providers. Concerning quality performance, L&M found “there appears to be little relationship between savings and high or low CAHPS scores.”
In sum, CMS staff could only conclude that results to date were “encouraging” largely because their assessments did not address “the degree to which the Pioneer ACO model can sustain small increases in spending and high quality performance over longer periods.” This left CMS staff to admit further, “ACOs may have faced challenges sustaining their performance.” Despite mixed results, CMS in April “certified” that the Pioneer demo reduces spending without harming quality, thereby allowing the Secretary to expand the demo or make it a permanent program.
Next Generation ACOs
This past March CMS announced the agency’s successor to the Pioneer ACO demonstration, functionally titled Next Generation. Those familiar with Pioneer’s regulations will immediately see the similarities in Next Generation; it is an amped up version of the Pioneer. Next Generation’s prospective beneficiary assignment, performance period, quality measures, and financial benchmarking will be similar to Pioneer. Savings will be determined, however, by discounting the financial benchmark by 0.5 to 4.5 percentage points via a composite score that combines quality performance and a regional and a national cost trend.
The real difference here is risk-reward is magnified. Next Generation providers can earn, or lose, substantially more money. While Pioneer offered up to 60 percent in shared savings, Next Generation will offer up to 80 percent for the first three years and 85 percent in years four and five, or participants can select a 100 percent risk track.
To incent risk, CMS will provide a number of enhancements, all strongly supported by MSSP participants. CMS will allow for multiple payment arrangements and improved beneficiary risk scoring. The agency will allow beneficiaries to voluntarily align or attest their participation in a Next Generation ACO, and it will give beneficiaries an opportunity to receive direct payments, up to $50 annually, for obtaining a certain percent of their care from a Next Generation provider. CMS will also make available to Next Generation providers the ability to utilize skilled nursing care without the beneficiary having had a prior hospitalization. The agency will allow expanded reimbursement coverage for telehealth and also allow for specific post-discharge home visits.
Despite greater opportunity for savings along with numerous operating, performance, and payment enhancements, CMS expects to award no more than 15 and 20 Next Generation contracts over the next two years.
Medicare Shared Savings Program ACOs
Of the current 405 ACOs, 400 providers participate in the so called no risk Track 1. (“So called” because ACOs, particularly physician-lead ACOs, invest a substantial amount of their own money in creating and managing their ACO. “No risk” because they can share financial savings but not owe money they’re reimbursed beyond their benchmark.) While performance year two results are expected this September, here’s what we know about performance year one (PY1) or 2012-2013. Of the 215 Track 1 ACOs operating in PY1, 52 earned shared savings. Similar to Pioneers, success was concentrated. Twelve ACOs were responsible for half of all savings.
Total savings were, however, questionable. Had PY1 not been pay-for-reporting but instead pay-for-quality, the $300 million the 52 ACOs collectively earned in shared savings would have been reduced by approximately 25 percent due to imperfect quality scores. In fact, 60 percent of the successful 52 ACOs had quality scores below the mean quality score for all 215 Track 1 ACOs, raising the question: “What is the correlation between quality measurement and better health outcomes? It is seldom if ever noted that 42 of the 215 Track 1 ACOs so exceeded their financial benchmark that, if they had chosen to participate in the at-risk Track 2, they would have collectively owed CMS $232 million, or roughly the same amount the 52 ACOs earning shared savings would have been paid had, again, PY1 been pay for quality. Five PY1 Track 2 ACOs achieved results similar to the Pioneers: two earned shared savings; two neither earned savings nor owed money; and one owed losses.
Since 98 percent of the MSSP program participants are in Track 1 and since less than a quarter of these earned shared savings in PY1, to what extent did CMS restructure the program via its June 4 final rule to improve the prospects for increased Medicare savings?
CMS did extend Track 1 for a second, three-year contract period at the same 50 percent earned shared savings, not a reduced 40 percent as the agency proposed. To address the problem of expected diminishing returns in subsequent contract periods, for those that earn shared savings CMS will add the ACO’s portion of shared savings to its reset benchmark should the ACO provider sign a subsequent contract. CMS will also equally weigh the three benchmark years when resetting an ACO’s benchmark in subsequent contract periods. It is expected this will result in more generous reset benchmarks and reduce the possibility of gaming performance during one contract period to obtain a higher or better financial benchmark in the next contract period.
While these improvements were all strongly supported by the provider community, CMS chose not to act on a larger number of reforms the ACO provider community equally desired. CMS created a new MSSP Track 3 that would reward upwards of 75 percent in earned shared savings; CMS will also assign Track 3 beneficiaries, as in the Pioneer demo, prospectively. This is helpful because it limits the problem of beneficiary churn or unstable assignment, whereby year-over-year 20 percent or more of beneficiaries are either leaving or entering an ACO, making it difficult if not impossible to manage patient care at an aggregate level. CMS will also advantage Track 3 ACOs by again waiving the three-day hospital stay requirement before utilizing SNF services and will consider further granting additional home health, telehealth, and post acute referral waivers to Track 3.
For both risk tracks 2 and 3, CMS will offer beneficiary attestation, whereby the beneficiary voluntarily attests participation with their ACO, thereby reducing patient churn and growing the number of an ACO’s participants. For Track 2 and 3 CMS will also allow the ACO a choice of medical savings rates that allows them to receive first-dollar savings for taking on greater reimbursement risk.
Prospective assignment, payment waivers, attestation, minimum savings rate relief, and additional improvements to risk adjustment and quality performance benchmarking were all reforms the ACO provider community argued for Track 1. In addition, the ACO provider community uniformly advocated for including a regional cost factor in calculating ACO benchmarks. CMS ruled it will incorporate regional costs in future rulemaking. However, this methodological change will not be implemented until 2017 and will not be available until 2019 for ACOs beginning in 2016 — the vast majority of whom will likely again choose Track 1.
From a practical standpoint, the final rule accomplished little. CMS simply recognized the necessity of a second Track 1 contract term and made two needed second term benchmark modifications. CMS failed to take the logical next step and provide Track 1 participants with performance enhancements such as prospective assignment and beneficiary attestation. Instead, the agency chose to double down by creating a second risk track and offering payment waivers and other enhancements to the risk tracks only.
That CMS would withhold improvements from Track 1 participants, when again they constitute essentially the entire program, is troubling. The agency recognizes Track 1 providers need more time to achieve success. The agency admits physicians, like people generally, are unwilling to risk loss even in the face of commensurate gains. Achieving shared savings or avoiding losses is therefore important if not necessary. CMS also knows the agency needs to grow the MSSP program substantially over the next few years because the Department is largely relying on ACO participants to achieve its goal of tying 50 percent of traditional Medicare payments to alternative payment models.
Lastly, CMS knows it now needs to prepare providers to compete on performance beginning in 2019 under the MACRA-created Merit-Based Incentive Payment System (MIPS) or qualify as a pay for value Alternative Payment Model provider. Knowing all this, how does CMS expect 90 percent of first-year Track 1 ACOs to renew their contracts later this year when it has not adequately given these providers a glide path to risk contracting?
What Else Is Known?
We also know related evidence does not prove how paying for value is achieved. Results from the Physician Group Practice (PGP) demonstration, upon which the MSSP is based, found only half of the ten participating large multispecialty groups generated savings and one of these earned more than half of the total savings amount. Moreover, PGP providers might have saved money regardless of the demo, since the practices that saved money during the demonstration had achieved similar savings pre-demonstration.
In the first year of the CMS Comprehensive Primary Care (CPC) demonstration, participants did not generate net savings for the demo’s attributed Medicare beneficiaries despite the fact that the demo includes a per-member per-month care management fee and provides national and regional learning support. Findings to date from CMS’s bundled payment demonstration are equally indeterminate. The demo’s evaluator, the Lewin Group, could only conclude this past February, “We are limited in our ability to draw conclusions,” and a recently published study in Health Affairs showed there was no difference in spending between bundling hospitals and a comparative hospital group.
The frequently referenced Alternative Quality Contract (AQC) program in Massachusetts did not produce net savings until its fourth year, and these savings are questionable because AQC provider performance may have been influenced by spillover effects from other near market global budget contracts. The AQC program is voluntary, therefore selection bias may have been in play. Regardless, the extent to which the AQC program is generalizable to the Medicare program is questionable: The average age of an AQC patient is 35, and unlike the Medicare population it’s certain one-third of AQC patients do not suffer a cognitive impairment.
At least four related points are worth noting. Over 50 percent of PY1 ACO providers were physician-led. This is relevant because physician practices, as opposed to insurance companies, lack the technical skill or ability to undertake, in accordance with state laws, the numerous tasks associated with evaluating and managing risk contracts — for example,
- defining a scope of capitated services;
- determining an adequate patient-panel size and appropriate case mix;
- calculating financial risk that includes accounting;
- providing for reinsurance, stop loss and risk corridors, and a risk reserve capital base;
- accounting for practitioner-to-practitioner variability;
- and addressing related administrative support services and costs.
CMS’s movement to risk also takes little or no account of the primary care physician’s (PCP) perspective. Polling data frequently finds physicians generally and PCP’s particularly are unwilling to be accountable or at risk for the cost of care — that substantially explains PCP movement to concierge medicine. Our own polling shows over and again the vast majority of our members are currently unwilling to take on risk contracting, no more so than those ACOs whose contracts expire this year.
Thirdly, there appears to be little concern about the well-recognized flaw within fee-for-service (FFS) Medicare, recently highlighted by a May GAO report, that primary care services are comparatively under-valued. It’s been long argued that reimbursement levels should be increased for primary care cognitive work and reduced for specialist procedural work. If CMS simply rolls up FFS payments, it may just be reinforcing or compounding this underlying problem.
Lastly, research over the past decade shows physicians can be or are ambivalent about financial incentives. For example, a study also recently published in Health Affairs found quality-incented PCPs in Minnesota did not out- perform a comparable non-incented group.
What Are We Forced To Conclude?
We wholeheartedly support CMS’s efforts to move the Medicare program to value-based payment and we are working diligently to support our members’ efforts to lower Medicare cost growth. At the same time, we believe, because of limited and mixed performance evidence to date, it would be well worth CMS’s time to allow practice to confirm theory. That is, allow performance evidence to accumulate over the next few years both in the Medicare program and in pay-for-value commercial markets.
Meanwhile, CMS should redouble its efforts to support actual ACO providers and near-term MSSP participants. Payment reform can be a means to improve care quality, patient outcomes and cost efficiency, but these goals should not be attempted via motivated reasoning.
No comments:
Post a Comment