Toward Accountable Payment

Demystifying the Mechanics of Bundled Payments

by Rob Lazerow

Learn about the payment mechanics of emerging bundled payment models, including the key differences among pricing bundles, retrospective reconciliation, and gainsharing within a bundled payment program.

Despite the recent release of the Medicare Shared Savings Program final rule, Advisory Board members continue to ask a range of provocative questions about CMS’ Bundled Payments for Care Improvement Initiative. Based on my recent conversations with providers across the country, we are clearly approaching a new wave of payment innovation projects as various organizations consider experimenting with performance risk and utilization risk.

Recently, many of members’ bundled payment questions have focused on the payment mechanics of a bundling program. For example, members have recently asked us:

  • What exactly would we be discounting if we participate in the bundled payment initiative?
  • How would we determine what savings to include in our physician gainsharing model?
  • How would the retrospective bundling process used for Models 2 and 3 impact gainsharing?
  • Would we gainshare around the bonus we get if we come in under our target during retrospective reconciliation in Model 2 or 3?

Combining these questions together, the overall question I want to address in this post is: how do discounting, retrospective reconciliation, and gainsharing all fit together in a bundled payment program?

The short answer

The high-level answer is that discounting, retrospective reconciliation, and gainsharing are all distinct parts of the mechanics of bundled payment:

  • Discounting is part of the process providers would use to establish their new reimbursement rates for the cases being bundled.
  • Retrospective bundling is the method providers would use to collect their bundled revenue under Models 2 and 3.
  • Gainsharing is part of the process hospitals would use to distribute funds out of the bundled payment once collected, rewarding physicians and potentially post-acute care (PAC) providers for successful performance improvement.

Taken together, discounting and retrospective reconciliation are both central to establishing a bundling program’s top line revenue, while gainsharing is ultimately a function of reducing the program’s internal input costs. Collectively, these factors all combine to determine a bundling program’s bottom line profitability.

The detailed answer

In order to provide a bit more nuance about these three related but distinct parts of bundled payment mechanics, I want to walk through an example of providers bundling total joint replacement under Model 2 of the Bundled Payments for Care Improvement Initiative. In the explanation below, I have separated bundled payment funds flow into the following three parts:

  • Determining the bundled reimbursement rate
  • Collecting the bundled revenue
  • Distributing the bundled payment

Determining the bundled reimbursement rate

When applying for the bundled payment initiative, applicants would begin by defining and pricing bundles for their selected admissions. Generally speaking, the applicant would add up all of the components of care related to the admission, combining together utilization rates and standard Medicare reimbursement rates for the hospital, surgeon, anesthesiologist, radiologist, any consulting physicians, PAC providers, and related readmissions. Taken together, this sum equals the amount that Medicare is currently reimbursing providers for the full episode of care through individual payments; we will call this amount the baseline bundled rate. For our sample bundle, we will assume that the baseline bundled rate price is $10,000 for the joint replacement case.

Next, the applicant would need to discount the baseline bundle according to the minimum discount rates set forth in the Request for Application (RFA). According to the RFA, the minimum discount rate for Model 2 depends on the duration of the bundle after the patient is discharged from the hospital; bundles spanning 30 to 89 days post-discharge have a minimum discount rate of 3%, while bundles spanning 90 or more days post-discharge have a minimum discount rate of 2%. For our sample bundle, we will set the duration post-discharge at the minimum 30 days, accepting the higher discount rate in exchange for limiting the time frame of the episode.

Taking our baseline bundled rate of $10,000 from above and apply the 3% minimum discount rate, the discounted bundled reimbursement rate becomes $9,700. The $300 difference between the baseline and discounted bundled rate represents Medicare’s full savings for the bundled case.

At this point, the applicant has successfully defined and priced the bundle, establishing the new reimbursement rate per bundled case.

Collecting the bundled revenue

Once the bundled payment initiative launches, providers participating in Models 2 and 3 will collect their bundled revenue though a retrospective bundling model. Under this “virtual” bundling methodology, all providers—hospitals, physicians, and PAC providers—would continue to receive their individual fee-for-service payments at their standard Medicare reimbursement rates. After an episode concludes, Medicare would complete a reconciliation process, calculating the total reimbursement paid out to providers for the episode and comparing that aggregate amount to the discounted bundle price established above. If the aggregate provider reimbursement amount were less than the bundled rate, Medicare would pay the difference to the providers. If, however, providers were reimbursed more than the stated bundled rate, they would need to repay the overage amount to Medicare.

For our joint replacement example, we will assume that providers successfully make the episode of care more efficient to Medicare by reducing some unnecessary outpatient imaging and the amount of post-acute care needed, collectively saving Medicare a total of $500 through these avoided services. Assuming the providers otherwise follow the historical trends used when pricing the bundle, they would have individually billed Medicare a total of $9,500. During the retrospective reconciliation process, Medicare would compare this actual spending amount to the discounted bundled rate of $9,700, and then pay the bundling program an additional $200 to reconcile the difference.

Although the retrospective bundling model includes a form of bonus payment—in this case the extra $200 the program receives after coming in under the discounted bundled rate—this payment is totally unrelated to the gainsharing process outlined below. The exclusive purpose of retrospective reconciliation is to simulate a true prospective bundled payment by creating a framework for paying providers the bundled rate without requiring that they actually accept and distribute a single lump sum payment. Under a prospective bundling model, conversely, the providers would have received a single $9,700 payment at the outset, and then accrued the financial benefits of improving efficiency.

Ultimately, the retrospective reconciliation process is used to collect bundled reimbursement, establishing a bundling program’s top line revenue.

Distributing the bundled payment

Finally, after providers collect the bundled revenue, they must determine how to distribute the bundled payment. In Model 2, this is straightforward, as the participating providers have already received their individual, non-discounted payments from Medicare through the retrospective reconciliation process detailed above. Under Model 4, however, the hospital would need to develop a process for distributing the prospective bundle, including paying physicians their professional fees—but that is a topic worthy of a future blog post.

Under Model 2, bundling programs have just one key consideration related to distributing funds: how to maximize the gainsharing opportunity enabled by the bundling initiative. After all, earning legal sanction to gainshare with other providers, especially independent physicians, is one of main benefits of participation. Regardless of the bundling model in question, providers will need gainsharing to motivate cost reduction efforts in order to recoup the price cuts outlined above. Across the inpatient stay, prime input cost reduction targets may include physician preference items, basic supplies, length of stay, ancillary services, and physician consults. In the context of an episodic bundled payment, such as Model 2, we would consider post-acute care costs and readmissions as additional input costs.

As part of designing effective gainsharing models, hospitals would need to calculate their input cost baseline to establish a starting point for later measuring cost savings. For many hospitals, this would require renewed focus on cost accounting to ensure accuracy. Once the bundling program is up and running, the hospital would determine how much a bonus-eligible physician or PAC provider has reduced input costs by comparing their actual input cost to the baseline. In addition to demonstrated cost savings, effective gainsharing models can feature a range of levers to drive performance improvement.

Returning to our illustrative example, we will assume that the baseline input cost for the joint replacement case is $6,000; as a starting place, the program’s margin on the baseline bundle (i.e., before discounting) would have been 40%. After successful cost reduction efforts, we will assume that the participating physicians and PAC providers could reduce input costs by $1,000. Next, given the design of the gainsharing model, the program would pay half of the demonstrated savings to partner providers as bonuses, ultimately retaining $500 of the actual savings. As a result, the effective input costs of the case would be down to $5,500. Combining these improved input costs with the $9,700 bundled reimbursement rate, the new margin on the bundled case, after paying out performance-based bonuses, would be approximately 43%.

Ideally, the cost savings fueled by gainsharing will more than make up for the initial price discount, helping hospitals maintain—or even improve—bottom line profitability.

Learn from successful bundling programs

To learn more about the mechanics of bundled payment programs, read the Health Care Advisory Board’s Succeeding Under Bundled Payments publication and the Financial Leadership Council’s "Finance Road Map for Bundled Payment" whitepaper. As always, please feel free to email me with your bundled payment questions.