A Payment Reform Conundrum: Reconciling Conflicting Policy Goals

In early 2015, the Secretary of the Department of Health and Human Services (HHS), Sylvia Burwell, announced goals for shifting Medicare payment from fee-for-service (FFS) to alternative payment models (APMs). Qualifying APMs include, among others, accountable care organizations (ACOs), medical homes, and episode-based bundled payments.

Congress, by large bipartisan majorities, endorsed payment reform in the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA), which establishes incentives for physicians to join new payment models. Many private purchasers, insurers, and providers are likewise committed to payment reform.

So far, the gains from alternative payment models have not been large. In 2014, 27 percent of ACOs in the Medicare Shared Savings Program shared savings, and the factor most strongly associated with success was having a history of high costs (and thus a more lenient financial benchmark). Slightly more than half of Pioneer ACOs shared savings, which may be attributable to stronger financial incentives or their greater experience with managed care (or some combination of the two).

Despite these results, there are reasons to be hopeful. ACOs appear to be improving quality, and they seem to become more effective with experience. The relative success of the Pioneer model may indicate that stronger incentives could spark better performance in reducing costs. Most importantly, the Centers for Medicare and Medicaid Services (CMS), private payers, and providers are investing time, energy, and money in refining the payment models.

Conflicting Goals

Some vexing choices lurk just ahead, however. Advocates of payment reform have goals that conflict with one another. Three goals are central to a "payment reform conundrum." The goals are:

1. To encourage providers to volunteer to participate in APMs;
2. To encourage providers in APMs to bear the risk of a financial loss, not just share in savings; and,
3. To reward provider efficiency, not just improvement relative to the provider's own past performance.

Encourage Providers To Volunteer For APMs

The new payment models may promise improvements in quality and efficiency, but provider participation is usually voluntary. So long as that remains the case, providers will be tempted to keep one foot on the sturdy and familiar "dock" of FFS while the other foot is tentatively placed in the "boat" of an APM.

Should the APM boat be deemed too risky, the provider may dive back to the safety of the dock. The onus is on CMS, therefore, to make APMs more attractive to providers than FFS. That will be difficult to accomplish while still saving Medicare money.

Encourage Providers To Bear The Risk Of A Financial Loss As Well As Sharing In Savings

So-called "two-sided models" create stronger incentives. (Behavioral economists have found that fear of a loss is more motivating than the potential for a gain.) However, requiring APM providers to bear risk may discourage them from participating.

CMS recently announced that 22 out of 434 Medicare Shared Savings Program ACOs have volunteered for Tracks 2 or 3, under which they bear financial risk. An additional 30 ACOs are in the risk-bearing Next Generation or Pioneer models. In the fourth year of the ACO effort, then, only 11 percent of ACOs have volunteered to accept risk.

Reward Provider Efficiency, Not Just Improvement Relative To A Provider's Own Past Performance

Support for payment reform rests, in large part, on its potential to reduce health care costs. Paying the largest rewards to providers with the lowest costs (holding quality constant) would, therefore, seem to be a cardinal principle. Yet ACO financial benchmarks are initially based on provider-specific historical costs, which means that providers beginning with high costs may earn larger bonuses than providers entering the program with low costs. Setting benchmarks to reward low costs, not improvement relative to a given's provider's past performance, may align the ACO model with the long-term goal of rewarding efficiency, but it could cause high-cost providers to abandon payment reform for the safety of FFS.

In short, Secretary Burwell's targets emphasize Goal 1 (increasing voluntary provider participation in APMs), but Goals 2 and 3 are important if APMs are to fulfill their potential. On the other hand, Goals 2 and 3 are threatening to many providers, especially providers with high costs, which could then reduce participation.

Because of the tension among the goals, policymakers face some difficult questions: Is it more important to get as many providers as possible into APMs or to strengthen the incentives for improving care? If some providers are likely to opt out, is it more important to get high-cost providers into the new models, or to reward already efficient providers for their success? Should FFS fees be cut to drive providers into APMs, or should policymakers focus on positive inducements to lure providers into them?

Resolving The Conundrum

There is no one solution to this conundrum. The likelihood of a satisfactory result will be increased, however, if policymakers look for options beyond the particulars of the APMs themselves. Changes in FFS and Medicare Advantage may be as important as APM design. One can imagine four policy "dials" that might be turned to higher "settings" than currently:

1. Make APMs mandatory
2. Make APMs and risk bearing more attractive
3. Make FFS less attractive
4. Strengthen incentives for Medicare beneficiaries to seek out low-cost/high-value providers

Make APMs Mandatory

Over the last several decades, Medicare has implemented many new payment systems, while compelling providers to participate. However, earlier reforms have been specific to particular provider types — for example, the inpatient hospital payment system covers services provided by the hospital during an admission but not physician services or post-acute care. CMS is now implementing APMs that bridge provider "silos," compelling different types of providers to collaborate. Bridging silos is overdue, but forging new relationships can be challenging.

Requiring providers to participate may be easier for episode-based bundles, as opposed to ACOs, because the range of services covered, and thus the number of new collaborators, is more limited. Indeed, CMS is compelling participation in its new joint replacement bundle, at least in certain markets. If that test is successful, one could imagine that approach being extended to additional cases.

Make APMs And Risk Bearing More Attractive

New models, and refinements of earlier editions, have cascaded out of CMS: three Medicare Shared Savings tracks, Pioneer ACOs, Next Generation ACOs, the ACO Advanced Payment and Investment Models, and a Comprehensive ESRD Model, not to mention episode bundles and new models for primary care. CMS is searching for designs and features that will entice providers not only to participate but also to assume financial risk.

CMS is attempting to make risk bearing more attractive for ACOs by increasing the providers' share of gains, reducing the Minimum Savings Rate (the threshold that must be crossed before an ACO is eligible to share savings), exempting risk-bearing ACOs from certain regulations, increasing ACO opportunities for beneficiary engagement, and shifting the method of payment from FFS towards capitation. These changes are appropriate and well-designed, but many ACOs are likely to remain risk averse.

Pioneer ACOs bear financial risk, but by Year 3 many had quit the program, opting instead for MSSP Track 1 (no risk); a return to FFS; or perhaps choosing to focus their efforts on Medicare Advantage. Some resigning Pioneer ACOs complained that the financial benchmarks favored providers with high costs and that opportunities for engaging with beneficiaries were weak, making it too difficult to get beneficiaries to seek care from within the ACO's network. CMS has responded with the Next Generation model of ACO, which includes elements addressing these and other issues.

The Next Generation model illustrates the conundrum, however: Shifting financial benchmarks to reward efficiency over improvement may keep some ACOs in the game, or even encourage them to assume risk, but it will likely have the opposite effect on other ACOs. Perhaps CMS can find the "golden mean" between efficiency and improvement that will encourage all ACOs to accept the Next Generation model, but it does not seem likely. Twenty-one organizations have enlisted in the Next Generation, but only nine Pioneers remain. In the two programs combined, then, there are 30 ACOs, two fewer than signed up for the Pioneer model in 2012.

Make FFS Less Attractive

MACRA sets higher payment rates for physician services provided through an alternative payment model by tightly constraining FFS payments and by paying a temporary 5 percent bonus on APM services. With this fee structure, the law pushes providers away from FFS, complementing CMS' efforts to lure them into the new payment models.

MACRA's 5 percent bonus for services provided through an alternative payment model does not take effect until 2019, and it expires at the end of 2024. For 2026 and subsequent years, payment rates for APM services would increase by 0.75 percent per year, but only by 0.25 percent for services paid under FFS. It is unclear how effective these payment differences will be in encouraging APM participation, but the gap could be increased in future legislation. MACRA's effect would also be amplified if Congress applied the payment differential to all providers, not just physicians.

Strengthen Incentives For Medicare Beneficiaries To Seek Out Low-Cost/High-Value Providers

Beneficiaries assigned to APMs retain their right to seek services from any Medicare provider. This limits APMs' potential to shift volume from poor-performing providers to high performers. Shifts in market share, and the resulting competition among providers, can induce providers to improve their performance just as APMs strive to do.

Medicare Advantage has greater potential to shift volume than APMs, since beneficiaries are rewarded for accepting limited provider networks through a combination of lower premiums and expanded benefits. About one-third of seniors are enrolled in Medicare Advantage, almost twice as many as are assigned to all of the ACO models combined.

Congress could amend Medicare Advantage to create stronger incentives for plans to reduce costs and for beneficiaries to seek out low-cost options. For example, the government's contribution toward coverage could be based on the average bid submitted by private plans and by traditional Medicare. Beneficiaries opting to enroll in plans with higher than average costs, including traditional Medicare, would pay more. (Currently, a beneficiary choosing traditional Medicare pays a fixed premium set by statute regardless how high Medicare costs are relative to private plans.) Such a model would bring Medicare into closer alignment with the operation of exchanges under the Affordable Care Act and recent changes in many employer-sponsored benefit plans.

Secretary Burwell's targets focus on Goal 1, getting providers to participate in APMs. To concurrently pursue Goals 2 and 3—getting providers to assume risk and rewarding efficiency—policymakers will need to turn all four "dials" described here. All four are currently turned "on," but, as outlined above, the settings are not high enough.

CMS is doing its best to make APMs more attractive, but Congress should do more. How far Congress turns the third and fourth dials—making FFS less attractive and providing stronger incentives to beneficiaries—will determine whether needed progress is made on all three goals.

from Health Affairs Blog http://ift.tt/1QD87Oy