Seeing “IHA” and “Fails” together in the title of an article in the nation’s premier health policy journal was not an outcome that we anticipated when the bundled payment initiative described by M. Susan Ridgely et. al in the August issue of Health Affairs was launched.

The key objective of the Integrated Healthcare Association (IHA)’s initiative was to implement over 20 payer-provider bundled payment contracts, resulting in completion of more than 500 bundled cases within the first two years of the project. During the third year of the project, researchers were to conduct both a clinical and an economic evaluation to test how bundled payments affect the quality and cost of care, in conjunction with an implementation evaluation to determine the scalability of this approach.

Looking back, these targets seem highly optimistic; but at the pilot’s launch, both IHA and its stakeholders had a number of reasons to be confident. The pilot was well funded by a three-year grant from the Agency for Health Research and Quality (AHRQ), building on two rounds of planning and feasibility work over four years funded by the Blue Shield of California Foundation and the California HealthCare Foundation. In addition, there was a high level of interest and enthusiasm among a core group of providers and health plans that had a prior history of collaboration in a California physician pay for performance program.

As we approached the end of the second year of the pilot, only a few contracts had been executed between plans and providers, yielding a small number of bundled cases completed. Although some progress was being made on additional contracts, it was clear that we would not achieve the quantitative goals we had envisioned. We shared this status update with AHRQ, which had funded our demonstration and evaluation.

Recognizing that we were unlikely to execute the targeted number of contracts within the grant period, AHRQ suggested that we focus our efforts on another component of the grant-funded activity: supporting the researchers conducting the implementation evaluation, fully documenting all that could be learned from this experience, and broadly disseminating the results. In that way, other payers and providers could benefit from our experience and make more rapid progress on their own bundled payment initiatives.

While it is difficult to acknowledge and accept failure, particularly in a widely-read and well-respected public forum such as Health Affairs, “learning from failure” must be more than just a platitude in the high stakes of health reform. To be clear, we agree that the process and outcome of this pilot were accurately described by the evaluators. However, it is important that the results be interpreted not as a death sentence for bundled payment, but as an early, hard-fought battle offering important lessons.

As the implementation team, we were very interested in the evaluation’s “Lessons for Future Bundled Payment Programs” that could be shared with others – and somewhat disappointed with the brevity of the findings. Project staff and partners worked diligently on each of the issues identified by the evaluators (e.g., “identify technical solutions for administering bundled payment”); the issues, efforts undertaken, and results are described in detail in Wesley Kary’s excellent white paper.

The problem isn’t that we didn’t anticipate these somewhat obvious issues; the problem is that they are very hard to solve in a collaborative way that works for all of the partners in a voluntary initiative. Indeed, then-IHA staffer Emma Dolan’s blog post describes the real-time challenges faced by the project team as the demonstration entered its second year. In the spirit of providing specific guidance to those interested in pursuing this pathway, we offer a few hard-won recommendations.

Start Retrospective, Then Transition to Prospective

A key implementation decision, not highlighted by the evaluators, is whether to adopt a prospective versus a retrospective payment model for bundled payment. The IHA initiative chose the former — prospective payment is well established in California, and stakeholders involved in the pilot felt that a retrospective payment approach (in which fee for service claims would continue to flow and be reconciled later to determine shared savings) would not fully test the impact of bundled payment.

Any bundled payment model requires considerable effort by clinicians and hospital administrators to redesign the approach to care delivery, but a prospective payment model also faces off against existing legal, regulatory, insurance benefit design and claims payment systems in ways we did not fully appreciate.

Prospective bundled payment raised numerous concerns for California regulators charged with protecting consumers, including whether providers were assuming insurance risk, how existing copayments and coinsurance would be applied, and whether consumers should be made aware of the payment arrangement. Nor did we anticipate the difficulty commercial payers would face adjudicating prospective bundled payments.

Based on our experience, we recommend that those interested in testing bundled payment consider initially testing the bundled payment model using a retrospective payment approach. As reported in a recent issue brief by HCI3, retrospective payment is the most common approach currently in use, easing the regulatory and administrative burdens in the early going. It also offers the advantage of developing a reliable financial baseline from which a prospective payment amount can be fairly negotiated. Importantly, this recommendation should not be interpreted as backing away from prospective bundled payment as the ultimate goal; rather, it is a practical, transitional step.

Put Care Redesign at the Center of Bundled Payment Implementation

Bundled payment is generally touted as a promising example of payment innovation — but the true benefit of bundling payments derives from reengineering care delivery, not from combining separately paid line items into a single tab. Bundled payment provides the impetus, but the work of care redesign must follow if the promise of bundled payment is to be realized: reductions in unnecessary care, reductions in readmissions, lower risk and complication rates for patients, and improved patient function and outcomes.

This important work can only be successful with strong clinical leadership backed by committed management. This formula was effectively demonstrated by the Hoag Orthopedic Institute. Hoag was an active participant in the bundled payment demonstration that undertook a major initiative to redesign care for total hip and knee replacement aimed at both improving care and ensuring that bundled payments would cover the costs of these procedures.

Care redesign requires significant attention, and can easily be overwhelmed by the myriad of other administrative details necessary to implement bundled payment. It is with this challenge in mind that, in its own Bundled Payment for Care Improvement (BPCI) initiative, Center for Medicare and Medicaid Services (CMS) has put equal emphasis on the care redesign and administrative aspects of bundled payment.

Target Markets in Which Bundled Payment Represents a Significant Step Forward

It may be tempting to look at the results reported by Ridgely et. al and conclude that if a group of experienced innovators can’t make bundled payment work in California, where prospective payment is historically embedded, its prospects seem dim elsewhere. In reality, the deep experience in the California market with prospective payments — capitation in particular — worked against the pilot.

In order for HMOs to participate in the initiative, they would have had to unwind existing capitation arrangements to allow for prospective bundled payments. This essentially limited participation in the demonstration to PPO contracts. In addition, California providers and health plans were deeply engaged in the development of Accountable Care Organizations (ACOs) and related contractual arrangements. Total population management, the holy grail of the ACO movement, was a more logical extension of existing capitation payment arrangements. To some organizations, particularly non-hospital based physician organizations, bundled payment appeared to be a distraction.

In many areas of the U.S., where the health care delivery system is highly fragmented and reimbursement remains largely on a fee-for-service basis, bundled payment offers a more immediate on-ramp to value-based payment. For example, the state of Arkansas tackled retrospective bundled payment as its initial foray into value based payment based in part on a relatively fragmented delivery system with low risk-bearing among provider systems.

Follow the Leader: CMS

One of the IHA demonstration’s core limitations should be familiar to anyone who has ever championed a voluntary initiative: it requires agreement from multiple partners, each with a small slice of plan membership and patient population. This underscores the importance of Medicare as a driver of bundled payment, based on its large enrollment and high utilization rates – particularly in procedures such as joint replacements, a favorite target of bundled payment. Commercial payers often follow Medicare’s lead, and strong leadership from CMS is essential to catalyze the investment of commercial payers in large-scale bundled payment implementation.

Fortunately, CMS has clearly indicated that bundled payment is an important element of its value-based purchasing strategy. This much is clear from CMS’s implementation of the ACE Demonstration Project and the more recent and larger-scale Bundled Payments for Care Improvement. Further evidence of this leadership is the under-the-radar national implementation of bundled payment by CMS for dialysis. This initiative has offered extensive results, both positive (lower costs, decline in heart attacks and strokes) and negative (higher blood transfusions initially at certain locations), as recently articulated by Jonathan Blum as he departed CMS.

For bundled payment to take hold on a national scale, CMS must stay the course with these initial efforts and ultimately develop a standard set of bundled payments, ideally prospective. This will take time – and there will certainly be other initiatives with disappointing results along the way.

Owning Our Failure – And Our Contribution

For those of us who have worked on this initiative, the sting of failure has long since worn off, replaced by a commitment to sharing both our tangible materials and our insights.

Among them, the demonstration project:

  1. Catalyzed payers and several established technology vendors to develop and test payment software to accommodate bundles, inspiring at least one national health plan to test and implement new software for this purpose.
  2. Activated interest among California provider organizations and health plans in clinical integration and value-based payment strategies.
  3. Prompted state regulators in California to articulate their concerns with prospective bundled payment, which in turn helped others understand and anticipate regulatory issues in their respective states.
  4. Produced ten well-vetted and fully-specified bundles, available to serve as a starting point for other innovators attempting to define and implement bundled payment.
  5. Created specifications and methods to analyze retrospective cost data to understand episode costs and price bundled payments.
  6. Produced contracting models and templates which have been used by payers in their own bundled payment initiatives.
  7. Engaged leading provider organizations in the important and difficult process of care redesign.

Notwithstanding the failure label, we believe the demonstration was a success in some significant respects. Contracts were signed, and patients seen; providers were paid prospectively for clearly-defined episodes of care, and this activity has continued beyond the end of the demonstration project. In spite of the challenges, a policy roundtable that we convened at the close of the initiative generated clear consensus among participants that the effort was worthwhile – and that payers, providers, and policy makers should continue to pursue bundled payment.