At the recent Becker’s Hospital Review Annual Meeting in Chicago, several presentations covered accountable care organizations, bundled payments and other new payment methodologies. Although not described specifically in these presentations, two different payment arrangements are emerging in which payers and providers are cooperating to reduce healthcare costs. This article compares and contrasts those payment methodologies.
Different contract arrangements
Two different payment models are arising in these types of contracts. Each has its own financial incentives and disincentives, as well as advantages or disadvantages to the providers and payers. The objectives of these payment methodologies are generally to align the financial incentives of the providers (hospitals, physicians, post-acute, etc.) with those of the payers (government, private insurers, employers, etc.) in a way that reduces costs to the payer. This reduction in the payer’s cost, therefore, originates in decreases in payments to the provider community as a whole. The process by which this decrease is created can create significantly different incentives for the participants. The goal of the payer is to create an environment in which healthcare cost decrease, and to realize those decreases. Therefore, payment systems that do not provide a predictable, properly incentivized financial structure may be less effective at achieving the payers’ goals.
From the provider's point of view, though, an important goal should be to retain as much revenue within the provider system as possible, and to allocate those funds to provide more efficient healthcare services. Revenue lost to the provider system and reclaimed by the payer reduces the resources available to the provider to implement care management programs.
Fixed discount models
Some arrangements, including one described at the Becker’s conference, utilize a payment model in which providers agree to a fixed payment amount, or a discount from a fixed payment amount, for all services included in the agreement. This was the arrangement in one seminar, in which the hospital contracted with a Medicare Advantage plan to receive a fixed capitation rate to cover all services for a specified population. There was no "shared savings" component to this arrangement – the payer presumably negotiated a sufficiently favorable capitation rate, and was content to allow the hospital and its physicians (which were collectively effectively functioning as an accountable care organization) to retain all of the revenue remaining in the capitation payment amount. The providers could allocate these funds as they chose, but those funds remain constant, and no further payments were made back to the payer.
This arrangement is similar to that used by Medicare in the Bundled Payment for Care Improvement (BPCI) initiative. In this program, CMS computes a fixed payment per episode based on its historical cost, decreases that payment by 2% or 3% depending on the length of the episode, and establishes the remaining amount as a fixed payment target for the episode. As in the situation described above, BPCI participants can allocate this payment in ways that they believe will create more efficient care, regardless of whether those services are currently paid by Medicare. For example, a bundled payment contractor could create and fund a post-discharge care team whose function was to reduce readmission rates. If this team was successful, readmissions would fall; however the hospital would not lose revenue from those decreased admissions because the episode payment rate would not decrease. There is no shared savings component – Medicare received its savings "off the top" and not as a result of any actions taken by the provider organization. The provider organization loses the amount of the discount, but retains all of the remaining revenue. Therefore, the provider organization and all of its component providers have no disincentive to reduce revenue-producing services, since those reductions will not actually result in lost revenue.
Shared savings arrangements
The other type of arrangement used in these emerging models is "shared savings" arrangement. In these arrangements, a fixed target payment amount for a specific purpose services (bundled payment episode or capitation payment amount for services provided to a population) is established. A contracting organization agrees to accept this budget as its revenue source, against which all payments for services provided to current beneficiaries are charged. If the cost of those services is below the revenue budget, the contracting organization receives a percentage of that "savings", with the payer retaining the remainder of that difference.
Under these arrangements, any reduction in provider services results in a decrease in provider revenue. This occurs because a portion of the revenue reduction (the “shared savings”) is shared with the payer. These funds depart the provider community and are not available for patient care, care management services or to create financial incentives for intra-provider cooperation.
Note that this decrease in provider revenue occurs for the provider community as a whole, but not necessarily for individual providers or groups thereof. For example, services provided by primary care physicians aren’t generally targets as a source of savings; indeed those services are generally increased in ACO arrangements. Therefore, PCPs rarely have disincentives to decrease utilization of services, since their own services are rarely at risk. Hospital services, though, are the most frequent source of shared savings through reductions in readmissions and emergency department services, and a shared savings arrangement can create a significant financial disincentive for hospitals to participate in service reductions.
The act of creating shared savings results in a decrease in revenue to some provider somewhere in the care delivery system. This is in contrast to the fixed discount arrangement in which the loss of provider revenue was created within the payer contract and is not affected by the subsequent actions of the provider. In essence, that discount is a “sunk cost” and creates no further incentives or disincentives for provider actions.
Comparisons of contract arrangements
In a fixed discount arrangement, the financial incentives of all providers are aligned. Since creating "savings" does not reduce the pool of payments available to providers, no provider is inherently disadvantaged by a reduction in services.
Shared savings arrangements have also been found to be less effective in creating savings for payers. In the report from the Congressional Budget Office entitled “Lessons from Medicare’s Demonstration Projects on Disease Management, Care Coordination and Value-Based Payment” , the CBO noted that the only Medicare demonstration project yielding savings at the time of the report was the heart bypass demonstration episode of care project in which CMS has negotiated a fixed percentage discount from the participants. Other programs in which savings to CMS were dependent on the ability of the contracting organization to achieve reductions in provider utilization were generally unsuccessful at achieving those goals.
Note that the form of the contract (fixed discount or shared savings) doesn’t affect the amount of savings that can be realized. Fixed discount contracts that have high discount amounts may transfer higher amounts of provider revenue to the payer than a shared savings model would, particularly if savings amounts are low. By contrast, an organization that creates significant “savings” through successful care management activities will lose a significant portion (in a Medicare ACO it’s about 50%) of that provider revenue, and that amount might exceed the discount that might be offered in a similar arrangement. The actual amount of payment depends on the details of the arrangement, not on its structure.
An advantage to the shared savings arrangement is that no funds are lost to the provider community as a whole unless shared savings are actually achieved. This is in contrast to the fixed discount arrangement, in which the provider community loses revenue regardless of achievement of savings. The opposite situation is true for the payer, who receives a guaranteed amount of provider revenue regardless of the actions of the providers.
Summary and Conclusions
Two different models are emerging for new payment systems. In fixed discount arrangements, contracting organizations agree to a fixed payment for services provided, which may include a discount from historical or expected payment rates. Alternatively, in a shared savings model reductions in payments to providers that result from care management activities are shared between the contracting organization and the payer.
These models differ in two major features: predictability and financial incentives to providers. Fixed discount models offer predictability about the amount of provider revenue transferred to the payer; the payer is assured of receiving that discount amount regardless of the activities of the providers. Similarly, the providers know the amount of revenue that they will lose to the payer as a result of the arrangement.
The financial incentives also differ between the models. In a shared savings model, providers know that some of the revenue that’s reduced through care management activities will go to the payer, and leave the provider group (which includes the contracting organization or ACO). This may dissuade them from taking the steps to develop the “shared savings” that these organizations were established to create. By contrast, in a fixed discount model the payer’s share of revenue is established by contract and isn’t subject to the providers’ actions, and therefore the providers don’t have a disincentive to create better care models since the “savings” that they create remains within the provider system. Each model has its advantages, and both should be evaluated by providers and payers who are considering more advanced payment systems.
 An "episode" begins with an inpatient admission in a specific DRG, and includes all services provided to that patient for between 30 and 90 days after hospital discharge.