In January of 2013, Centers for Medicare and Medicaid Services launched its Bundled Payments for Care Improvement (BPCI)
initiative. This three-year pilot project will test four models in which Medicare will pay fixed sums for defined episodes of patient care.
Providers that accept bundled payments stand to be rewarded for quality and efficiency. However, they also accept risk for outliers, adverse events, readmissions and other complications that drive up costs.
This hasn't stopped over a hundred hospitals and post-acute facilities
from signing up to participate. Their enthusiasm underscores industry-wide concern about the coming shift from fee-for-service to value-based reimbursement.
However, one year into the initiative, many important questions remain. CMS has yet to reveal a detailed blueprint for compensation under its bundled payment program. So what can hospitals expect going forward? And is there anything to gain by watching and waiting?
BPCI (Somewhat) Demystified
Fundamentally, a bundled payment is a fixed payment intended to cover a defined episode of care. In acute care settings, the payment may be made to facilities, physicians, or jointly to both.
The overriding goal of bundled payments is to reduce costs by creating a payment system that rewards quality over quantity. To thrive under fixed payments, hospitals, physicians and post-acute providers must align their goals. They must create a strong Acute Care Continuum that minimizes fragmentation and maximizes coordination and continuity.
So how exactly will that happen? Well, that's what BPCI intends to find out. While Medicare and other public and private payers have experimented with bundled payments for decades, this will be the program's first large-scale trial.
BPCI will test four reimbursement models
- Model 1. Retrospective (adjusted) payments for hospital stays covering Medicare Part A services. Paid for patients of all DRGs.
- Model 2. Retrospective payments for hospital stays PLUS a defined period of post-acute care covering Medicare Part A and B services. Paid for 48 defined "episodes of care" based around clinical conditions.
- Model 3. Retrospective payments for post-acute care only covering Medicare Part A and B services. Paid for 48 episodes of care.
- Model 4. Prospective (non-adjusted) payments for hospital stays covering Medicare Part A and B services. Paid for 48 episodes of care.
While Medicare has provided a broad framework for the initiative, uncertainties remain. So how worried should potential participants be about them?
Up in the Air
Here are the most pressing bundled payments questions as I see them:1. Are hospitals ready to implement the necessary changes?
To succeed with bundled payments, providers will need to take a hard look at the long-term effects of care decisions. They will need data to support cost analysis and protocols and mechanisms to prevent unnecessary and duplicative care. Perhaps most fundamentally, they will need to redesign care pathways in order to align incentives and foster continuity and coordination.
This is undoubtedly a tall order. However, results of smaller-scale risk-sharing experiments suggest it's far from impossible. A notable example is a 2010 pilot program in which several hospitals in rural Maryland were placed on fixed operating budgets. By 2013, the hospitals had increased their capacity to care for all but the most acute patients in outpatient settings. They also invested in home health and social work programs to boost continuity and prevent costly readmissions. Within a few years the hospitals were reporting strong operating profits
, and patients were reporting greater satisfaction with their care.2. Does BPCI represent a true end point?
Looking at the models tested, it's probably fair to say that they're transitional rather than a preview of coming attractions. Three of the four calculate reimbursements on a "retrospective" basis, meaning actual expenditures for a DRG or episode of care are reconciled against a target price. To some degree, this still rewards providers for a high volume of services, which doesn't exactly move us toward a system that values quality over quantity.
So where are we really
headed? Suzanne Delbanco director of the nonprofit organization Catalyst for Payment Reform (CPR), believes retrospective payments are a transitional measure
, and that CMS' ultimate goal is to provide prospective payments only. If she's right, providers piloting the first three models should see them as a bridge to Model 4 — truly prospective (fixed-sum) payments that are determined before the fact and adjusted only in rare cases.3. How should fixed payments be divided?
This is perhaps the thorniest open question in the bundled payment puzzle. When a patient is rushed to the ED with acute myocardial infarction, what percentage of the bundled payment does the hospital get? The ED doctor? The hospitalist? The cardiologist or specialist? And who decides?
The Robert Wood Johnson Foundation and Nixon Peabody LLP have published an issue brief
that could serve as a starting point for providers grappling with these questions. The document includes a sample participating provider agreement that includes tables of definitions, price lists, and sample mechanisms for determining shared savings bonuses.
Of course, one of the goals of BPCI is to encourage innovation, so it's likely that in a couple years we'll have a better understanding of what works and what doesn't. But for the pilot participants, this could be a significant challenge.4. Will bundled payments actually save money?
BPCI is quite a leap when you consider that no large-scale studies have demonstrated the financial benefits of bundled payments. However, experiences of smaller programs do suggest some gains:
- In an early CMS demonstration involving bundled payments for coronary artery bypass graft (CABG), seven hospitals saved a total of about $42.3 million, or 10 percent of expected expenditures, over five years.
- A bundled payments initiative for knee surgeries between Blue Cross and Blue Shield of North Carolina and CaroMont Health saved an average of 8 to 10 percent per episode.
Can these limited findings be generalized to a program involving 48 episodes of care and facilities of all sizes, shapes and locations? The answer almost certainly varies by provider. However, it probably stands to reason that those volunteering for the program believe they have many of the building blocks for success in place.5. Will bundled payments improve quality?
Again, we can only judge by the experiences of more limited programs, but these do suggest cause for optimism. For example, Geisinger Health System's ProvenCare program charges a flat fee for elective CABG. By standardizing the surgery along best-practice guidelines, Geisinger significantly reduced complication rates
, resulting in a 67 percent drop in inpatient mortality and a 76 percent reduction in deep wound infections.
Participation is Optional, Preparation is Crucial
Given these uncertainties, should your organization join in the pilot program? I think it's a question that every provider needs to evaluate on an individual basis. A one-year "preparation" period is built into BPCI before risk sharing kicks in, but it may not be sufficient time for every group or facility.
On the other hand, risk sharing is coming as surely as the sun rises. So it's essential that providers take steps to prepare for payment reform whether or not they join a formal pilot program.
In many ways, the changes needed to succeed with bundled payments are the same quality goals that providers have been working toward for the past decade. They are valuable goals because they represent what is best for patients. Duplicative, unnecessary care and system fragmentation create very real hardships for the sick and injured. So despite the uncertainties inherent in BPCI, I do believe that payment reform will ultimately move acute care in a positive direction.
[Image credit: "Pumpkin-Pie-Whole-Slice"
licensed under CC BY-SA 3.0