Quantcast
Channel: Health Affairs Blog » Scott Heiser
Viewing all articles
Browse latest Browse all 2

Balancing Goals In The MSSP: Consider Variable Savings Rates

0
0

Blog_Austin

Having started with a solid design and attracted a sizable number of organizations, the Centers for Medicare and Medicaid Services (CMS) continues to improve and evolve the options for health systems to participate in value-based payment reform. On Thursday, June 4, CMS released a highly anticipated final rule for changes to the Medicare Shared Savings Program (MSSP), one of its accountable care organization (ACO) programs. The rule ratified numerous changes proposed in the December filing, including offering a new Track 3 for ACOs ready to take on greater risk and reward (75 percent savings or loss sharing rates) in exchange for more predictable prospective attribution and the option to receive first dollar savings. CMS also lowered the eligibility requirements and is allowing continuation in Track 1—with upside risk only—for another contract period to encourage ACOs to remain in the program.

A companion piece by our colleagues at the Brookings Institution and KPMG summarizes all of the major elements of the final rule. We focus on the steps CMS took (and has yet to take) to address ACOs’ concerns about how the financial targets, which determine whether an ACO is successful at saving or guilty of overspending, are determined.

Two Important Benchmarking Changes

Whether savings have been achieved is decided by comparing actual annual spending to a benchmark — the hypothetical level of spending that CMS predicts would have occurred absent payment reform. This is essentially a two-step process. Step one entails determining baseline spending for a given year. Step two involves determining how much should be added to the baseline to set benchmark spending for the following year(s).

In the final rule, CMS unveiled two important changes to the way baseline spending is calculated in a re-based contract in hopes of maintaining MSSP participation. First, CMS will change the way benchmarks are generated for ACOs continuing in the program. Both the original and current rule use a three-year weighted moving average to establish baseline spending. The original rule weighted the previous year 60 percent, the year before 30 percent, and the prior year 10 percent. While this weighting scheme will stay in place for new ACOs joining the MSSP, now each of the three prior years will be equally weighted for MSSPs staying in the program and re-signing 3-year contracts. The logic behind this change is that as ACOs accrue savings over time, the original equation would ratchet down benchmark spending more quickly for successful ACOs, perhaps to unachievable levels.

Second, CMS will add back into the new benchmark any savings that were generated by the ACO, thus increasing the benchmark by the saved amount, again so as not to penalize successful ACOs. Both of these changes will make it easier for ACOs in the MSSP to achieve savings.

Additional Benchmarking Decisions Delayed

CMS postponed decision-making about further changes to benchmarking, including how to incorporate regional or national information, and determining how much to add to baseline spending to determine the benchmark for future years. Rather, CMS indicated that it would initiate a new rulemaking process this summer to address the methodology they will use to establish benchmark spending levels. This will include not only how to determine the growth rates to apply to each ACO’s baseline spending to calculate a benchmark, but also how and when to “rebase” baseline levels. CMS suggests its rulemaking process will resolve whether ACOs continuing in the MSSP will continue to have baseline spending figures derived from historical spending of its population, derived from a regional comparison, or another method.

Setting these spending targets in a way that gives program participants confidence that real changes in performance will be rewarded is critical to the long-term sustainability of the program and the willingness of ACOs to re-sign for future agreements. The setting of the benchmark is of absolute importance; in the absence of “savings” achieved in comparison to the benchmark, the other contract characteristics, such as the minimum savings rate (the savings threshold ACOs must cross before sharing in savings), the sharing rate (the percentage of savings going to the ACO and to CMS), or quality performance (how ACO scores on quality metrics affect their eligibility to share savings), are irrelevant because they don’t come in to play.

In the proposed rule released in December, CMS discussed a broad array of options for revising how spending benchmarks are set. A detailed description of the potential approaches (and combinations of approaches) and their potential strengths and weaknesses are provided in the Federal Register. The challenge that CMS faces is to balance a number of competing interests, including: encouraging current ACOs to remain in the program; encouraging other providers to join; avoiding windfall gains that result from having benchmarks that favor high-cost or low-cost ACOs, or those that happen to be located in high- or low-cost growth regions; encouraging continued improvement in quality; and achieving meaningful savings for both Medicare beneficiaries and taxpayers.

CMS Could Consider a More Nuanced Approach

We believe that CMS may be paying insufficient attention to the possibility that a more nuanced way of setting shared savings rates might help balance the many competing interests at play.

Why might this be both important and helpful? First, ameliorating ACO concerns that the infrastructure costs needed to launch an ACO will be recouped is essential. Second, evidence has emerged that suggests substantial inequities in the current model that favor ACOs with higher baseline spending. Under the current benchmarking scheme, Pioneer ACOs with higher benchmark spending were more likely to earn savings and Medicare Shared Savings Program participants in Hospital Referral Regions in the highest quintiles of Medicare spending were more likely to achieve savings.

An illustration is provided by the entrepreneurial physicians in McAllen, TX, first identified in Atul Gawande’s 2009 piece “The Cost Conundrum,” who are continuing to thrive: apparently each physician in one of the McAllen ACOs received pay of $800,000 under the ACO model, even after accounting for overhead and technology investments, due to the benchmarking procedure. Physicians in McAllen were doing well before the emergence of ACOs; they should not receive windfall gains under ACO reforms because of their prior poor performance. High cost health care systems are high cost not as a consequence of natural forces (i.e. sicker patients), but because of the overuse of discretionary and potentially avoidable services. Systems with high baseline spending should not be rewarded through a methodology that makes it easier for them to achieve outsized savings.

Creating a sustainable ACO program will involve designing a conceptual model that is responsive to each of these concerns: covering the costs of infrastructure development in early years to encourage participation, and ensuring fair financial compensation that does not advantage higher-spending ACOs or those in higher-spending areas. A model that incorporates graduated savings distributions may help account for operational costs and historical inefficiencies.

The Rationale For Graduated Savings Distribution

Conceptually, it would be reasonable to provide all ACOs with a high degree of confidence that if they improved their performance they would benefit, such that there is an incentive to join the program and confidence in the prospect of easily recouped costs of investment. CMS might sharpen this incentive by giving ACOs a larger proportion of savings in the early years of participation, and reserving a larger share for the Medicare Trust Fund as savings grow over time. Further, the sharing rate would be adjusted to account for the size of initial benchmarks — higher spending ACOs at baseline could be assigned lower savings rates.

Modeling A Graduated Savings Distribution Approach

For illustrative purposes, imagine something like the following. In the first tranche below the benchmark, an ACO would keep all savings up to 4 percent. These sharing rates would decrease as savings increased in magnitude: between 4 – 6 percent of savings below benchmark, an ACO would keep 75 percent; between 6 – 10 percent of savings below benchmark, it would keep 50 percent. As savings climbed past, say, 10 percent (the 95th percentile of savings was 11 percent in 2012-2013, see Exhibit 1), an ACO would keep just 25 percent. (See Exhibits 2 and 3.)

First dollar savings at these rates would occur after the chosen minimum savings rate is met. Savings rates might be further raised as ACOs increase the proportion of their total population under ACO contracts (with Medicaid and private payers), or savings rates might vary depending on whether baseline spending was in higher or lower quartiles. There are numerous technical specifics to be detailed with such an approach, especially around how to set the best thresholds and savings rates, as well as the influence of the minimum savings rate, the risk and reward ratios across the ACO tracks, and how such details can be operationalized.

Simulations with current data on ACO performance would help inform whether this approach could address the concerns above. Using the above thresholds and savings rates for MSSP ACOs in 2012 – 2013, we modeled the potential financial gains and losses from such an approach for ACOs and for CMS. (See Exhibits 2 and 3.) Offering ACOs a larger share of savings below the benchmark (100 percent of savings below 4 percent, and 75 percent of savings between 4-6 percent) would be expected to generate $178.6 million in additional funds for ACOs. Conversely, giving CMS a larger share as savings climbed past a level indicative of windfall gains (10 percent in simulation) would grant the Medicare Trust Fund $25.8 million more than under current rules, though CMS would gain less across the program due to these changes. Our aim with this simulation is not to presume we have identified the “right” sharing rates or thresholds at which they might apply, but to illustrate the potential for improving equity in the program, among other goals.

Looking Forward

All of the evaluations of Medicare’s ACO programs to date suggest that the model has promise. The quality of care improves for beneficiaries. Savings are generated not only for some of the ACOs, but also for the Trust Fund (better than estimates from Medicare Advantage). In addition, there may be spillovers of these programs on non-Medicare populations resulting in improved quality and lower cost growth overall.

The challenges facing CMS include encouraging provider organizations to join and remain in the program while ensuring that at least some savings continue to flow to beneficiaries and American taxpayers. CMS introduced many creative and promising ideas in the proposed rule. Those included in the final rule released last week were generous to ACOs; the decisions still to be made appear more challenging. CMS has made it clear that there will not be an eternal path with only shared savings — participating providers will need to shift to downside risk as well.

Decisions regarding the benchmark would almost certainly benefit from timely simulations and transparency on the findings so that program participants—and those considering applying—could make more informed decisions about the likely benefits and risks. If we can get the incentives right, providers, patients and payers and purchasers (employers and tax-payers) could all share the benefits of a more integrated, higher-quality health care system.

Exhibit 1
Fisher_Exhibit1_MSSP-ACO-Benchmarkign

Exhibit 2

Fisher_Exhibit2_MSSP-ACO-Benchmarking

Source: Center for Medicare and Medicaid Services (CMS) 2012 – 2013 Medicare Shared Savings Program performance year results. Note: To calculate current savings returned to ACOs and CMS, we calculated shared savings for all ACOs that sufficiently reported quality and saved more than their Minimum Savings Rate (MSR); this was to counteract CMS recouping Advance Payments from savings of ACOs that received upfront payments. Simulated figures are calculated to reflect sharing in the absence of the Advance Payment program.

Exhibit 3

Fisher-Exhibit-3-MSSP-ACO-Benchmarking


Viewing all articles
Browse latest Browse all 2

Latest Images

Trending Articles





Latest Images